DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 02/05/2026 has been entered.
Response to Arguments
Applicant’s arguments, filed on 02/05/2026, with respect to the 35 USC § 101 rejections for claims 1-3, 6-10 and 12-23 have been considered but they are not persuasive (Arguments, pages 10-15).
The Applicant argues the rejection of independent claims 1, 13, and 19 under 35 U.S.C §101 by summarizing what the Office Action asserts, i.e. each claim is allegedly (1) directed to an abstract idea, and more specifically, to a mental process, (2) fails to incorporate the abstract idea into a practical application, and (3) fails to recite significantly more than the abstract idea. Then the Applicant submits amendments to independent claims 1, 13, and 19, and therefore, the Applicant asserts that the present response render the current rejections under 35 U.S.C §101 is moot with respect to those amended claims due to the idea that . The Applicant, for example, asserts that amended independent claim 13 recites involving a method of operating a natural language understanding (NLU) framework that cannot be practically performed as a mental process in the human mind. The method recited by amended independent claim 13 features "performing lookup source inference of different portions of an utterance simultaneously across the plurality of taxonomy lookup sources to determine a plurality of taxonomy segmentations for the utterance, wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources, and indicates corresponding alternative tokens for each of the matched tokens of the utterance." The Applicant asserts that even if the human mind is capable of simultaneous processing, parallel processing threads are an inherently technical approach to computer processing. The human mind is not arranged in software-based processing threads, and is, therefore, incapable of performing lookup source inference using parallel processing threads. Accordingly, claims 1, 13, and 19 have been amended to include steps that cannot practically be performed in the human mind or via a pen and paper. Thus, claims 1, 13, and 19 are valid at prong one of the patent eligibility framework. Thus, the Applicant asserts that a method of operating a natural language understanding (NLU) framework that involves performing lookup source inference of an utterance simultaneously across a plurality of taxonomy lookup sources to determine a plurality of taxonomy segmentations for the utterance cannot be practically performed as a mental process in the human mind. (Arguments, pages 10-13).
The Examiner respectfully disagrees. Though the Applicant amends independent claims 1, 13, and 19, with parallel processing threads and therefore, and as a result the Applicant asserts that the present response overcomes the §101 rejection. The Examiner respectfully disagrees that the Applicant’s assertion of amended independent claim 13, for instance, involving a method of operating a natural language understanding (NLU) framework that “cannot be practically performed as a mental process in the human mind” because of the amended features requiring of "performing lookup source inference of different portions of an utterance simultaneously across the plurality of taxonomy lookup sources to determine a plurality of taxonomy segmentations for the utterance.” The Applicant’s argument and assertion that improvement to meaning search accuracy and steps for improving a specific technical field with using parallel processing threads do not really integrate into a practical application. This is because parallel processing is a well-known and a well-established method and it can just be considered as an additional element. For instance, a patent application published in 2008 and authored by Lunenfeld Pat App No. US 20080021906 A1 (Lunenfeld) claims to employ parallel processing of multiple queries/keyword searches of multiple information sources of the same and/or different types and may be used on substantially any kind of network; quick response intelligence gathering of multiple same and/or different information requests of multiple sources, grouping and sorting results substantially simultaneously in real time and on-the-fly…and is particularly useful for corporate, industrial, commercial, and government purchasing of multiple products from multiple sources… presenting results to single and/or multiple users substantially simultaneously in real time and on-the-fly (Lunenfeld, para 0826-0831).
The Applicant further argues that claims 1, 13, and 19 provide an improvement to meaning search accuracy in a natural language understanding framework. In at least these ways, claims 1, 13, and 19 include steps for improving a specific technical field and, therefore, integrate any alleged judicial exception into a practical application (Arguments, page 14).
The Examiner respectfully disagrees. Though the Applicant amends independent claims 1, 13, and 19, with parallel processing threads, and as a result the Applicant asserts that those amendments might produce improvement to meaning search accuracy in a natural language understanding framework to introduce steps for improving a specific technical field, the Examiner respectfully disagrees with that characterization. In fact, those specific aspects which the Applicant considered to be improvement are some of the aspects that are basically directed towards the abstract idea. The parallel processing added as the amendment is used as an “apply it” similar to generically applying a processor as in Lunenfeld (para 0826-0831).
Similarly, as detailed in the next section, claims 2-3, 6-10, 12, 14-23 recite similar claim language as in claims 1, 13 and 19, and also add some more abstract ideas using mathematical concepts, and therefore claims 2-3, 6-10, 12, 14-23 are rejected as abstract ideas as well.
Thus, claims 1-3, 6-10 and 12-23 are still rejected under 35 U.S.C. § 101.
Applicant’s arguments, filed on 02/05/2026, with respect to the 35 USC § 103 rejections for claims 1-3, 6-10 and 12-23 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument(Arguments, pages 15-19).
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-3, 6-10 and 12-23 are rejected under 35 U.S.C. §101 because the claims recite an abstract idea and mental process. Claims 1, 13 and 19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The independent claims 1, 13 and 19 recite “lookup source system …; performing… lookup source inference; and performing, … vocabulary injection to generate a plurality of re-expressions of the utterance based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens indicated by a particular taxonomy segmentation with the corresponding alternative tokens indicated by the particular taxonomy segmentation …” as drafted cover an abstract idea of data analysis/retrieval and mental steps. More specifically, the “lookup source system has a plurality of taxonomy lookup sources that each includes a respective source data representation compiled from taxonomy source data that represents one or more hierarchal relationships between product categories within a domain of a client, wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories; and at least one processor configured to execute stored instructions to cause the NLU framework to perform actions comprising: performing, via the lookup source system, lookup source inference of different portions of an utterance simultaneously across the plurality of taxonomy lookup sources to determine a plurality of taxonomy segmentations for the utterance, wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources, and indicates corresponding alternative tokens for each of the matched tokens of the utterance; and performing, via the NLU system, vocabulary injection to generate a plurality of re-expressions of the utterance based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens indicated by a particular taxonomy segmentation with the corresponding alternative tokens indicated by the particular taxonomy segmentation,” require just data analysis / retrieval step by gathering/acquiring lookup source system data …, performing inference and looking up source inference, as well as performing vocabulary injection to generate a plurality of re-expressions of the utterance require gather and analyzing data and also some of these can be done using pen and paper and some of these can be done mentally as well. The claimed invention is, therefore, directed to an abstract idea and a mental process without significantly more and thus, claims 1 and 11 are rejected under 35 U.S.C. 101.
Similarly, claims 2-12, 15-18 and 20-23 recite similar claim language as in claims 1, 13 and 19, and even add more abstract ideas using mathematical concepts. Claim 2 recites “taxonomy source data indicates vertical relationships, lateral relationships, or any combination thereof, between organizational entities within a domain of the client.” More specifically, “the taxonomy source data indicates vertical relationships, lateral relationships, or any combination thereof, between organizational entities within a domain of the client,” which requires just data analysis / retrieval step and a mental process that can be performed using pen/pencil and paper for indicating the taxonomy source data for vertical relationships, lateral relationships, or any combination thereof, between organizational entities within a domain of the client. Thus, claim 2 is directed to an abstract idea.
Claim 3 which recites “the taxonomy source data comprises a table having a plurality of vocabulary source columns and a hypernym column and wherein the at least one memory is configured to store a lookup source framework, and the at least one processor is configured to execute stored instructions to cause the lookup source framework to perform actions comprising: grouping the plurality of vocabulary source columns of the taxonomy source data using the hypernyms column as a pivot column to generate hypernym-grouped taxonomy source data table, wherein each entry the hypernym-grouped taxonomy source data table comprises a unique hypernym token and a set of alternative tokens related to the unique hypernym token; and compiling the respective source data representation of each of the plurality of taxonomy lookup sources from a particular entry in the hypernym-grouped taxonomy source data table, wherein each respective source data representation comprises a plurality of states, wherein the plurality of states includes: a plurality of original states, each having a respective state value representing the unique hypernym token or one of the set of alternative tokens of the particular entry; and a plurality of produced states, each having a respective state value derived by a producer of the lookup source framework from the unique hypernym token or one of the set of alternative tokens of the particular entry,” which also requires just data analysis / retrieval step and mental process. For instance, based on available data collected by pen and paper, one can create a grouping of the plurality of vocabulary source columns of the taxonomy source data using the hypernyms column as a pivot column to generate hypernym-grouped taxonomy source data table. In such created table, each entry of the hypernym-grouped taxonomy source data table can comprise of a unique hypernym token and a set of alternative tokens related to the unique hypernym token. Thus, claim 3 is directed to an abstract idea.
Claim 6 recites “before performing the lookup source inference of the utterance, parsing, via the NLU system, the utterance to generate a parsed utterance with part-of-speech (POS) tagging indicating nouns and noun-phrases of the utterance, wherein the lookup source system is configured to perform the lookup source inference only for the nouns and the noun-phrases of the parsed utterance to determine the plurality of taxonomy segmentations of the parsed utterance,” which require a mental process. Thus, claim 6 is directed to an abstract idea.
Claim 7 recites “the utterance is a received user utterance, and at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions comprising: generating, via the NLU system, an utterance meaning model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance; and performing, via the NLU system, a meaning search to extract an intent and entity of the utterance using the utterance meaning model,” which require just data analysis / retrieval step which can be performed using an ipad, a tablet, or at most with the use of conventional/generic (general-purpose) computer. Thus, claim 7 is directed to an abstract idea.
Claim 8 recites “wherein the utterance is a sample utterance of an intent-entity model, and at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions comprising: generating, via the NLU system, an understanding model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance; and performing, via the NLU system, a meaning search to extract an intent and entity of a received user utterance using the understanding model,” which require just data analysis / retrieval step and mathematical calculations. Thus, claim 8 is directed to an abstract idea.
Claim 9 recites “the NLU system comprises an intent-entity model, and the at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions comprising: before performing the lookup source inference of the utterance, receiving the utterance from a user as a newly submitted sample utterance to be validated for inclusion in the intent- entity model; after performing the vocabulary injection to generate the plurality of re-expressions of the utterance, providing, to the user, an indication that the newly submitted sample utterance is valid or invalid based on the plurality of taxonomy segmentations; and providing, to the user, the plurality of re-expressions of the utterance as suggestions of alternative valid sample utterances for inclusion in the intent-entity model,” which require just data gathering step and a mental step which can be performed by a human or using pen and paper. Performing the lookup source inference of the utterance, receiving the utterance from a user as a newly submitted sample utterance to be validated for inclusion in the intent- entity model, performing the vocabulary injection to generate the plurality of re-expressions of the utterance, providing the utterance as suggestions of alternative valid sample utterances for inclusion in the intent-entity model can be done mentally. Thus, claim 9 is directed to an abstract idea.
Claim 10 recites “to perform vocabulary injection, the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions comprising: for each taxonomy segmentation of the plurality of taxonomy segmentations of the utterance: for each matched token of the matched tokens of the utterance indicated by the taxonomy segmentation: selecting, based on a configuration of the NLU framework, a corresponding alternative token from the corresponding alternative tokens for the matched token indicated by the taxonomy segmentation , wherein the configuration defines a level of hypernym-based replacement; and generating one of the plurality of re-expressions of the utterance that substitutes the matched token of the utterance with the corresponding alternative token, and wherein to select the corresponding alternative token, the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions comprising: in response to determining that the corresponding alternative token complies with the level of hypernym-based replacement, selecting the corresponding alternative token from the corresponding alternative tokens indicated by the taxonomy segmentation,” which require just data analysis / retrieval step and mental step. Performing vocabulary… for each taxonomy segmentation of the plurality of taxonomy segmentations of the utterance…for each matched token of the matched tokens of the utterance indicated by the taxonomy segmentation, selecting a corresponding alternative token from the corresponding alternative tokens for the matched token and generating one of the plurality of re-expressions of the utterance can be performed mentally. Thus, claim 10 is directed to an abstract idea.
Claim 12 recites “the configuration indicates substitution of the matched token with the corresponding alternative tokens that are hypernyms, or synonyms, or formal names, or colloquial names of the matched token, and wherein, to select the corresponding alternative token, the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions comprising: in response to determining that the corresponding alternative token is a hypernym, or a synonym, or a formal name, or a colloquial name of the matched token, in accordance with the configuration, selecting the corresponding alternative token from the corresponding alternative tokens indicated by the taxonomy segmentation,” which require just data analysis / retrieval step and mental step. Substitution of the matched token with the corresponding alternative tokens, to select the corresponding alternative token, and selecting the corresponding alternative token can be done mentally. Thus, claim 12 is directed to an abstract idea.
Claim 14 recites “the taxonomy source data comprises a table having a plurality of vocabulary source columns and a hypernym column, and the method comprises: grouping the plurality of vocabulary source columns of the taxonomy source data using the hypernyms column as a pivot column to generate hypernym-grouped taxonomy source data table, wherein each entry the hypernym-grouped taxonomy source data table comprises a unique hypernym token and a set of alternative tokens related to the unique hypernym token; and compiling the respective source data representation of each of the plurality of taxonomy lookup sources from a particular entry in the hypernym-grouped taxonomy source data table,” which require just data analysis / retrieval step and mental step. Grouping the plurality of vocabulary source columns of the taxonomy source data and compiling the respective source data can be done mentally. Thus, claim 14 is directed to an abstract idea.
Claim 15 recites “the utterance is a received user utterance, and the method comprises: generating an utterance meaning model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance; and performing a meaning search to extract an intent and entity of the utterance using the utterance meaning model,” which require just data analysis / retrieval step and mental step. Generating an utterance meaning model and performing a meaning search to extract an intent and entity can be done mentally. Thus, claim 15 is directed to an abstract idea.
Claim 16 recites “the utterance is a sample utterance of an intent-entity model of the NLU framework, and the method comprise: generating an understanding model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance; and performing a meaning search to extract an intent and entity of a received user utterance using the understanding model,” which require just data analysis / retrieval step and mental step. Generating an understanding and extract an intent and entity can be done mentally. Thus, claim 16 is directed to an abstract idea.
Claim 17 recites “performing the lookup source inference of the utterance, receiving the utterance from a user as a newly submitted sample utterance to be validated for inclusion in the intent- entity model; after performing the vocabulary injection to generate the plurality of re-expressions of the utterance, providing, to the user, an indication that the newly submitted sample utterance is valid or invalid based on the plurality of taxonomy segmentations; and providing, to the user, the plurality of re-expressions of the utterance as suggestions of alternative valid sample utterances for inclusion in the intent-entity model,” which require just data analysis / retrieval step and mental step. Performing the lookup source inference of the utterance and receiving the utterance can be done mentally. Thus, claim 17 is directed to an abstract idea.
Claim 18 recites “performing vocabulary injection comprises: for each taxonomy segmentation of the plurality of taxonomy segmentations: for each matched token of the utterance indicated by the taxonomy segmentation: for each corresponding alternative token indicated by the taxonomy segmentation: generating one of the plurality of re-expressions of the utterance that substitutes the matched token of the utterance with the corresponding alternative token,” which require just data analysis / retrieval step and mental step. acquiring face image, face recognition and identification steps can be performed mentally or face image can be taken using a simple camera and one can recognize those images mentally. Thus, claim 18 is directed to an abstract idea.
Claim 20 recites “the utterance is a received user utterance, a sample utterance of an intent-entity model of the NLU framework, or a newly submitted sample utterance to be validated for inclusion in the intent-entity model,” which require just data analysis / retrieval step and mental step. Receiving user utterance… submitting sample utterance and validation can be done mentally. Thus, claim 20 is directed to an abstract idea.
Claim 21 recites “the hypernyms, hyponyms, or both of the product categories are specific to the domain of the client,” which require just data analysis / retrieval step and mental step. The hypernyms, hyponyms, or both of the product categories are specific to the domain of the client can be done mentally. Thus, claim 21 is directed to an abstract idea.
Claim 22 recites “the taxonomy segmentations, the alternative tokens, or both are specific to the domain of the client,” which require just data analysis / retrieval step and mental step. The taxonomy segmentations, the alternative tokens, or both are specific to the domain of the client can be done mentally. Thus, claim 22 is directed to an abstract idea.
Claim 23 recites “the hierarchal relationships indicate synonyms, formal names, or colloquial names of the product categories that are specific to the domain of the client,” which require just data analysis / retrieval step and mental step. The hierarchal relationships indicate synonyms, formal names, or colloquial names of the product categories can be done mentally. Thus, claim 23 is directed to an abstract idea.
Thus, claims 1-3, 6-10 and 12-23 as drafted cover a mental process, data gathering/retrieval and analysis/processing and/or mathematical abstract ideas, and they are mental processes directed to an abstract idea of implementing mathematical formulae for data processing and data analysis using a conventional/generic (general-purpose) computer as well (Spec. para 0067- 0069, and 0005) and thus, those claims are directed to an abstract idea.
This judicial exception is not integrated into a practical application. In particular, most of claims 1-3, 6-10, 12 recite additional element of “processor” and claim 19 recites “computer-readable medium,” and most of the claims 1-3, 6-10 and 12-23 recite “natural language understanding (NLU) framework” as per the independent claims, “parallel processing” as recited by independent claims 1, 13 and 19. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. This can simply be performed using a conventional/generic (general-purpose) computer. The claim is directed to an abstract idea.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to the integration of the abstract idea into a practical application, the additional element of using a computer is noted as a general computer as noted. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Further, the additional limitation in the claims noted above are directed towards insignificant solution activity. All of the claims can be performed using a calculator, an ipad, a tablet, or some other electronic device, or at most with the use of conventional/generic (general-purpose) computer (Spec. para 0067- 0069, and 0005). The parallel processing threads added as the amendment have been used as an “apply it” similar to generically applying a processor as in the prior art Lunenfeld (para 0826-0831) and therefore, claims 1-3, 6-10 and 12-23 do not contain patent eligible subject matter that has been identified by the courts.
Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional general purpose computer implementation. Claims 1-3, 6-10 and 12-23, are therefore not drawn to patent eligible subject matter as they are directed to an abstract idea without significantly more.
Therefore, the claimed invention is directed to an abstract idea and a mental process without significantly more and thus, claims 1-3, 6-10 and 12-23 are rejected under 35 U.S.C. 101.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2, 6-9, 13, 15-17, 19-24 are rejected under 35 U.S.C. 103 as being unpatentable over Galitsky Pub. No. US 10839154 B2 (herein after called Galitsky) in view of Sapugay et al. Pub. No. US 2021/0004537 A1 (herein after called Sapugay), further in view of Keysar et al. Pat App No. US 20220058232 A1 (Keysar), further in view of Yamada et al., "Hypernym discovery based on distributional similarity and hierarchical structures." In Proceedings of the 2009 conference on empirical methods in natural language processing, pp. 929-937. 2009 (Yamada), and further in view of Lunenfeld Pat App No. US 20080021906 A1 (Lunenfeld).
Regarding Claim 1, Galitsky discloses a natural language understanding (NLU) framework, comprising:
at least one memory configured to store a NLU system (Galitsky, para 539, system memory may store program instructions) and a lookup source system (Galitsky, para 263, database lookups), wherein the lookup source system has a plurality of taxonomy lookup sources (Galitsky, para 228, taxonomy-based) that each includes a respective source data representation compiled from taxonomy source data (Galitsky, para 228, taxonomy-based; para 420, source data); and
at least one processor configured to execute stored instructions to cause the NLU framework to perform actions (Galitsky, para 541, software (programs, code modules, instructions) that when executed by a processor provide the functionality described above may be stored in storage subsystem; i.e., at least one processor configured to execute stored instructions to cause the NLU framework to perform actions) comprising:
Galitsky does not specifically disclose performing, via the NLU system, vocabulary injection to generate a plurality of re- expressions of the utterance based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens indicated by a particular taxonomy segmentation with the corresponding alternative tokens indicated by the particular taxonomy segmentation.
However, Sapugay, in the same field of endeavor, teaches performing, via the NLU system, vocabulary injection (Sapugay, para 0094, performing vocabulary injection) to generate a plurality of re-expressions of the utterance ( Sapugay, para 0094, generate re-expressions of the utterances) based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens (Sapugay, para 0094, generate a new utterance in which the term “developer” is substituted by the term “employee”) indicated by a particular taxonomy segmentation with the corresponding alternative tokens ( Sapugay, para 0093; involve substituting certain tokens with other tokens) indicated by the particular taxonomy segmentation (Sapugay, para 0078; segmentations).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because vocabulary injection will enable one to introduce multiple re-renderings of the original utterance (Sapugay, para 0094).
Galitsky in view of Sapugay do not specifically disclose source data that represents one or more hierarchal relationships between product categories within a domain of a client.
However, Keysar, in the same field of endeavor, discloses disclose source data that represents one or more hierarchal relationships between product categories within a domain of a client, (Keysar, para 0023, Fig. 2, FIG. 2 shows an illustrative example of answering a question in accordance with some implementations of the present disclosure. In the illustrated example, a webpage search engine receives a “who” question, retrieves a number of search results, and provides an answer based on the search results and information associated with the search results. It will be understood that while the illustrated example refers to person entities in response to a “who” question, any suitable type of entity may be identified in response to any suitable type of question. For example, location entities may be identified in response to a “where” question; Keysar, 0032, In some implementations, the system calculates the sum the entity references occurring in list 230, list 232, and list 234, as shown in summation 236. In the illustrated example, there are 7 instances of [Juan Carlos I] identified in the top three search results shown in search result box 206. In some implementations, the system identifies [Juan Carlos I] as entity result 240 and provides answer 242 based on entity result 240. In some implementations, the system displays answer 242 including the text [The King of Spain is Juan Carlos I]. In some implementations, the system displays only text [Juan Carlos I]. In some implementations, the system generates a natural language or other format response based in part on the received query. In some implementations, the system may display additional information associated with the answer, for example, a picture of Juan Carlos I or a link to an encyclopedia entry. It will be understood that in some implementations, the system may display content in search results box 206 before displaying answer 240; Keysar, para 0049 - 0052, FIG. 4 shows illustrative knowledge graph 400 containing nodes and edges. Illustrative knowledge graph 400 includes nodes 402, 404, 406, and 408. Knowledge graph 400 includes edge 410 connecting node 402 and node 404. Knowledge graph 400 includes edge 412 connecting node 402 and node 406. Knowledge graph 400 includes edge 414 connecting node 404 and node 408. Knowledge graph 400 includes edge 416 and edge 418 connecting node 402 and node 408. Knowledge graph 400 includes edge 420 connecting node 408 to itself. Each aforementioned group of an edge and one or two distinct nodes may be referred to as a triple or 3-tuple. As illustrated, node 402 is directly connected by edges to three other nodes, while nodes 404 and 408 are directly connected by edges to two other nodes. Node 406 is connected by an edge to only one other node, and in some implementations, node 406 is referred to as a terminal node. As illustrated, nodes 402 and 408 are connected by two edges, indicating that the relationship between the nodes is defined by more than one property. As illustrated, node 408 is connected by edge 420 to itself, indicating that a node may relate to itself. While illustrative knowledge graph 400 contains edges that are not labeled as directional, it will be understood that each edge may be unidirectional or bidirectional. It will be understood that this example of a graph is merely an example and that any suitable size or arrangement of nodes and edges may be employed. Generally, nodes in a knowledge graph can be grouped into several categories. Nodes may represent entities, organizational data such as entity types and properties, literal values, and models of relationships between other nodes. A node of a knowledge graph may represent an entity, as defined above. In some implementations, entity types, properties, and other suitable content is created, defined, redefined, altered, or otherwise generated by any suitable technique. For example, content may be generated by manual user input, by automatic responses to user interactions, by importation of data from external sources, by any other suitable technique, or any combination thereof. For example, if a commonly searched-for term is not represented in the knowledge graph, one or more nodes representing that node may be added. In another example, a user may manually add information and organizational structures. A node representing organizational data may be included in a knowledge graph. These may be referred to herein as entity type nodes).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Keysar in the method of Galitsky in view of Sapugay because that would enable to rank the entity references and also to select any entity reference(s) from the ranking list for determination of an answer (Keysar, Abstract).
Galitsky in view of Sapugay and Keysar do not specifically disclose wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories.
However, However, Yamada, in the same field of endeavor discloses wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories (Yamada, p. 934, 1st col, Figure 4: [i.e., hierarchal relationships indicating the hypernym car and the hyponyms hybrid vehicle, mini vehicle, etc., which are also product categories]).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Yamada in the method of Galitsky in view of Sapugay and Keysar because that would extract a hierarchical structure relations that produces a hypernym of a certain word that may also be the hyponym of another hypernym with a deeper hierarchical structure method (Yamada, p. 931, 2nd col, 2nd para).
Galitsky in view of Sapugay, Keysar and Yamada do not specifically disclose performing, via the lookup source system, lookup source inference of different portions of an utterance across the plurality of taxonomy lookup sources using parallel processing threads to determine a plurality of taxonomy segmentations for the utterance.
However, Lunenfeld, in the same field of endeavor, discloses performing, via the lookup source system, lookup source inference of different portions of an utterance across the plurality of taxonomy lookup sources using parallel processing threads to determine a plurality of taxonomy segmentations for the utterance (Lunenfeld, Para 0826 – 0828, parallel processing of multiple queries/keyword searches of multiple information sources of the same and/or different types and may be used on substantially any kind of network; quick response intelligence gathering of multiple same and/or different information requests of multiple sources, grouping and sorting results substantially simultaneously in real time and on-the-fly; combined search and E-Commerce, and/or as a single point of purchase/sale for multiple products in multiple categories from multiple sites, and is particularly useful for corporate, industrial, commercial, and government purchasing of multiple products from multiple sources, as well as internet purchasing of multiple products from multiple sources )), wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources (Lunenfeld, Para 0829-0831, performing research, using multiple information sources, multiple sites, search engines, servers, databases, clients, applications, software applications, programs, and/or software programs, and may be performed in parallel using multiple queries/keyword phrases in multiple categories and/or multiple fields substantially simultaneously, in real time, and on-the-fly; downloading multiple title/subject and/or music/audio/video/television substantially simultaneously; presenting results to single and/or multiple users substantially simultaneously in real time and on-the-fly ).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Lunenfeld in the method of Galitsky in view of Sapugay, Keysar and Yamada because that would enable retrieving substantially multiple simultaneous services and/or information having the same and/or different criteria from the same and/or different servers, sorting, grouping, and/or organizing the responses from the servers and/or the clients into information and/or services responses, and communicating the service and/or information responses to the requesters and/or the users substantially simultaneously (Lunenfeld, para 0008).
Regarding Claim 2, Galitsky in view of Sapugay, and further in view of Keysar, further in view of Yamada, and further in view of Lunenfeld disclose all the limitations of claim 1 upon which claim 2 depends.
Galitsky further discloses the taxonomy source data indicates vertical relationships, lateral relationships, or any combination thereof, between organizational entities within a domain of the client (Galitsky para 228, taxonomy-based; para 107 and Figure 5: Figure 5 depicts text spans that are leaves, or terminal nodes, on the tree. The nodes indicate relationships; para 5 and Figure 5: Figure 5 depicts a node-link representation of the hierarchical binary tree; i.e., the taxonomy source data indicates vertical relationships, lateral relationships, hierarchical relationships, or any combination thereof, between entities within a domain of the client).
Regarding Claim 6, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of claim 1 upon which claim 6 depends.
Galitsky further discloses that at least one processor (Galitsky, para 534, one or more processors) is configured to execute stored instructions to cause the NLU framework to perform actions (Galitsky, para 539, system memory may store program instructions that are loadable and executable on processing unit) comprising:
Galitsky further discloses before performing the lookup source inference of the utterance, parsing (Galitsky, para 332, inferences; 263, database lookups; para 334, utterances ), via the NLU system, the utterance to generate a parsed utterance (Galitsky, para 76, discourse (i.e., utterance) parsers) with part-of-speech (POS) tagging (Galitsky, para 245, POS tags) indicating nouns and noun-phrases (para 171, noun phrase (NP), noun (N)) of the utterance, wherein the lookup source system (Galitsky, para 263, database lookups) is configured to perform the lookup source inference only for the nouns and the noun-phrases (Galitsky, para 171, noun phrase (NP), noun (N)) of the parsed utterance to determine the plurality of taxonomy segmentations of the parsed utterance (para 228, taxonomy-based; para 57, segments of discourse).
Regarding Claim 7, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of claim 1 upon which claim 7 depends.
Galitsky further discloses that the utterance is a received user utterance (Galitsky, para 116, user saying), and at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions (Galitsky, para 534, one or more processors; para 539, system memory may store program instructions that are loadable and executable on processing unit) comprising:
Galitsky also further discloses generating, via the NLU system, an utterance meaning (Galitsky, para 209, generating a discourse (i.e., utterance)) model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance (Galitsky, para 348, exchanges can consist of two to four utterances; para 203, the expressions that are the verbal definitions; the resultant expression).
However, Galitsky does not specifically teach performing, via the NLU system, a meaning search to extract an intent and entity of the utterance using the utterance meaning model.
Sapugay, in the same field of endeavor, further teaches performing, via the NLU system, a meaning search to extract an intent and entity of the utterance using the utterance meaning model (Sapugay, para 0010, a NLU framework and an intent-entity model having defined intents and entities that are associated with sample utterances. The NLU framework includes a meaning extraction subsystem that is designed to generate meaning representations for the sample utterances of the intent-entity model to construct an understanding mode).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because the disclosed NLU framework will be able to include a meaning search subsystem that is designed to search the meaning representations of the understanding model (also referred to as the search space) to locate matches for meaning representations of the utterance meaning model (also referred to as the search key) (Sapugay, para 0010).
Regarding Claim 8, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of claim 1 upon which claim 8 depends.
Sapugay, in the same field of endeavor, further teaches that the utterance is a sample utterance of an intent-entity model (Sapugay, para 0013, NLU framework that includes a meaning extraction subsystem capable of generating multiple meaning representations for utterances, including sample utterances in the intent-entity model and utterances received from a user), and at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions (Sapugay, para 0056, the one or more processors may include one or more microprocessors capable of performing instructions stored in the memory) comprising:
Sapugay also teaches generating, via the NLU system, an understanding model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance (Sapugay, para 0023, generates meaning representations from sample utterances of an understanding model to yield understanding model).
Sapugay also teaches further teaches performing, via the NLU system, a meaning search to extract an intent and entity of a received user utterance using the understanding model (Sapugay, para 0023, the meaning search subsystem compares meaning representations of the utterance meaning model to meaning representations of the understanding model to extract artifacts (e.g., intents and entities) from the received user utterance).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because this will enable for sample utterances of the intent-entity model to be used to generate meaning representations of an understanding model to define the search space for a meaning search (Sapugay, para 0061).
Regarding Claim 9, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of claim 1 upon which claim 9 depends.
Sapugay, in the same field of endeavor, further teaches the NLU system comprising an intent-entity model (Sapugay, para 0065, the NLU framework processes the utterance based on the intent-entity model), and the at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions (Sapugay, para 0056, the one or more processors may include one or more microprocessors capable of performing instructions stored in the memory).
Sapugay also teaches before performing the lookup source inference of the utterance receiving the utterance from a user, as a newly submitted sample utterance to be validated for inclusion in the intent-entity model (Sapugay, para 0010, the NLU framework includes a meaning extraction subsystem that is designed to generate meaning representations for the sample utterances of the intent-entity model to construct an understanding model, as well as generate meaning representations for a received user utterance to construct an utterance meaning model).
Sapugay also teaches after performing the vocabulary injection to generate the plurality of re-expressions of the utterance, providing, to the user (Sapugay, para 0094, block 308, performing vocabulary injection on the original utterance; Vocabulary injection generally involves introducing multiple re-renderings of the original utterance; para 0098, an utterance generated from vocabulary injection), an indication that the newly submitted sample utterance is valid or invalid based on the plurality of taxonomy segmentations (Sapugay, para 0078, Figure 7, the prosody subsystem of the meaning extraction subsystem analyzes the prosody of the utterance using a combination of rule-based and ML-based prosody plug-ins. Using these plug-ins, the prosody subsystem analyzes the utterance for prosodic cues, including written prosodic cues such as rhythm (e.g., chat rhythm, such as utterance bursts, segmentations).
Sapugay also teaches providing, to the user, the plurality of re-expressions of the utterance as suggestions of alternative valid sample utterances for inclusion in the intent-entity model (Sapugay, para 0092, Figure 11; the NLU framework generates re-expressions of an original utterance, and then generates a set of meaning representations based on these re-expressions and the original utterance. The original utterance may be one of the sample utterances of the intent-entity model).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because the vocabulary subsystem may access the vocabulary model of the understanding model to identify alternative vocabulary that can be used to generate re-expressions of the utterances having different tokens (Sapugay, para 94).
Regarding Claim 13, Galitsky discloses a method of operating a natural language understanding (NLU) framework, comprising:
Galitsky further discloses compiling a respective source data representation of a plurality of taxonomy lookup sources (Galitsky, para 263, database lookups) from taxonomy source data (para 228, taxonomy-based);
Galitsky does not specifically disclose performing, via the NLU system, vocabulary injection to generate a plurality of re- expressions of the utterance based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens indicated by a particular taxonomy segmentation with the corresponding alternative tokens indicated by the particular taxonomy segmentation.
However, Sapugay, in the same field of endeavor, teaches performing, via the NLU system, vocabulary injection (Sapugay, para 0094, performing vocabulary injection) to generate a plurality of re-expressions of the utterance (Sapugay, para 0094, generate re-expressions of the utterances) based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens (Sapugay, para 0094, generate a new utterance in which the term “developer” is substituted by the term “employee”) indicated by a particular taxonomy segmentation with the corresponding alternative tokens (Sapugay, para 0093; involve substituting certain tokens with other tokens) indicated by the particular taxonomy segmentation (Sapugay, para 0078; segmentations).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because vocabulary injection will enable one to introduce multiple re-renderings of the original utterance (Sapugay, para 0094).
Galitsky in view of Sapugay do not specifically disclose source data that represents one or more hierarchal relationships between product categories within a domain of a client, wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories.
However, Keysar, in the same field of endeavor, discloses source data that represents one or more hierarchal relationships between product categories within a domain of a client (Keysar, para 0023, Fig. 2, Fig. 2 shows an illustrative example of answering a question in accordance with some implementations of the present disclosure. In the illustrated example, a webpage search engine receives a “who” question, retrieves a number of search results, and provides an answer based on the search results and information associated with the search results. It will be understood that while the illustrated example refers to person entities in response to a “who” question, any suitable type of entity may be identified in response to any suitable type of question. For example, location entities may be identified in response to a “where” question; Keysar, 0032, In some implementations, the system calculates the sum the entity references occurring in list 230, list 232, and list 234, as shown in summation 236. In the illustrated example, there are 7 instances of [Juan Carlos I] identified in the top three search results shown in search result box 206. In some implementations, the system identifies [Juan Carlos I] as entity result 240 and provides answer 242 based on entity result 240. In some implementations, the system displays answer 242 including the text [The King of Spain is Juan Carlos I]. In some implementations, the system displays only text [Juan Carlos I]. In some implementations, the system generates a natural language or other format response based in part on the received query. In some implementations, the system may display additional information associated with the answer, for example, a picture of Juan Carlos I or a link to an encyclopedia entry. It will be understood that in some implementations, the system may display content in search results box 206 before displaying answer 240; Keysar, para 0049 - 0052, FIG. 4 shows illustrative knowledge graph 400 containing nodes and edges. Illustrative knowledge graph 400 includes nodes 402, 404, 406, and 408. Knowledge graph 400 includes edge 410 connecting node 402 and node 404. Knowledge graph 400 includes edge 412 connecting node 402 and node 406. Knowledge graph 400 includes edge 414 connecting node 404 and node 408. Knowledge graph 400 includes edge 416 and edge 418 connecting node 402 and node 408. Knowledge graph 400 includes edge 420 connecting node 408 to itself. Each aforementioned group of an edge and one or two distinct nodes may be referred to as a triple or 3-tuple. As illustrated, node 402 is directly connected by edges to three other nodes, while nodes 404 and 408 are directly connected by edges to two other nodes. Node 406 is connected by an edge to only one other node, and in some implementations, node 406 is referred to as a terminal node. As illustrated, nodes 402 and 408 are connected by two edges, indicating that the relationship between the nodes is defined by more than one property. As illustrated, node 408 is connected by edge 420 to itself, indicating that a node may relate to itself. While illustrative knowledge graph 400 contains edges that are not labeled as directional, it will be understood that each edge may be unidirectional or bidirectional. It will be understood that this example of a graph is merely an example and that any suitable size or arrangement of nodes and edges may be employed; Generally, nodes in a knowledge graph can be grouped into several categories. Nodes may represent entities, organizational data such as entity types and properties, literal values, and models of relationships between other nodes. A node of a knowledge graph may represent an entity, as defined above. In some implementations, entity types, properties, and other suitable content is created, defined, redefined, altered, or otherwise generated by any suitable technique. For example, content may be generated by manual user input, by automatic responses to user interactions, by importation of data from external sources, by any other suitable technique, or any combination thereof. For example, if a commonly searched-for term is not represented in the knowledge graph, one or more nodes representing that node may be added. In another example, a user may manually add information and organizational structures. A node representing organizational data may be included in a knowledge graph. These may be referred to herein as entity type nodes).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Keysar in the method of Galitsky in view of Sapugay because that would enable to rank the entity references and also to select any entity reference(s) from the ranking list in the determination of an answer (Keysar, Abstract).
Galitsky in view of Sapugay and Keysar do not specifically disclose wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories.
However, Yamada, in the same field of endeavor discloses wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories (Yamada, p. 934, 1st col, Figure 4: [i.e., hierarchal relationships indicating the hypernym car and the hyponyms hybrid vehicle, mini vehicle, etc., which are also product categories]).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Yamada in the method of Galitsky in view of Sapugay and Keysar because that would extract a hierarchical structure relations that produces a hypernym of a certain word that may also be the hyponym of another hypernym with a deeper hierarchical structure method (Yamada, p. 931, 2nd col, 2nd para).
Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld do not specifically disclose performing lookup source inference of different portions of an utterance across the plurality of taxonomy lookup sources using parallel processing threads to determine a plurality of taxonomy segmentations for the utterance, wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources, and indicates corresponding alternative tokens for each of the matched tokens of the utterance.
However, Lunenfeld, in the same field of endeavor, discloses performing lookup source inference of different portions of an utterance across the plurality of taxonomy lookup sources using parallel processing threads to determine a plurality of taxonomy segmentations for the utterance (Lunenfeld, Para 0826 – 0828, parallel processing of multiple queries/keyword searches of multiple information sources of the same and/or different types and may be used on substantially any kind of network; quick response intelligence gathering of multiple same and/or different information requests of multiple sources, grouping and sorting results substantially simultaneously in real time and on-the-fly; combined search and E-Commerce, and/or as a single point of purchase/sale for multiple products in multiple categories from multiple sites, and is particularly useful for corporate, industrial, commercial, and government purchasing of multiple products from multiple sources, as well as internet purchasing of multiple products from multiple sources ), wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources, and indicates corresponding alternative tokens for each of the matched tokens of the utterance (Lunenfeld, Para 0829-0831, performing research, using multiple information sources, multiple sites, search engines, servers, databases, clients, applications, software applications, programs, and/or software programs, and may be performed in parallel using multiple queries/keyword phrases in multiple categories and/or multiple fields substantially simultaneously, in real time, and on-the-fly; downloading multiple title/subject and/or music/audio/video/television substantially simultaneously; presenting results to single and/or multiple users substantially simultaneously in real time and on-the-fly ).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Lunenfeld in the method of Galitsky in view of Sapugay, Keysar and Yamada because that would enable retrieving substantially multiple simultaneous services and/or information having the same and/or different criteria from the same and/or different servers, sorting, grouping, and/or organizing the responses from the servers and/or the clients into information and/or services responses, and communicating the service and/or information responses to the requesters and/or the users substantially simultaneously (Lunenfeld, para 0008).
Regarding Claim 15, Galitsky in view of Sapugay, and further in view of Keysar, further in view of Yamada, and further in view of Lunenfeld disclose the method all the limitations of claim 13 upon which claim 15 depends, wherein the utterance is a received user utterance (Galitsky, para 330, an utterance).
Sapugay, in the same field of endeavor, teaches generating an utterance meaning model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance (Sapugay, para 0010, a NLU framework and an intent-entity model having defined intents and entities that are associated with sample utterances. The NLU framework includes a meaning extraction subsystem that is designed to generate meaning representations for the sample utterances of the intent-entity model to construct an understanding mode; para 0092, Figure 11; the NLU framework generates re-expressions of an original utterance, and then generates a set of meaning representations based on these re-expressions and the original utterance. The original utterance may be one of the sample utterances of the intent-entity model); and
Sapugay also teaches performing a meaning search to extract an intent and entity of the utterance using the utterance meaning model (Sapugay, para 0010, a NLU framework and an intent-entity model having defined intents and entities that are associated with sample utterances. The NLU framework includes a meaning extraction subsystem that is designed to generate meaning representations for the sample utterances of the intent-entity model to construct an understanding mode).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because the disclosed NLU framework will be able to include a meaning search subsystem that is designed to search the meaning representations of the understanding model (also referred to as the search space) to locate matches for meaning representations of the utterance meaning model (also referred to as the search key) (Sapugay, para 0010).
Regarding Claim 16, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the method all the limitations of claim 13 upon which claim 16 depends.
Sapugay, in the same field of endeavor, further teaches that the utterance is a sample utterance of an intent-entity model (Sapugay, para 0065, the NLU framework processes the utterance based on the intent-entity model) of the NLU framework.
Sapugay also teaches generating an understanding model that includes the utterance and at least a portion of the plurality of re-expressions of the utterance (Sapugay, para 0023, generates meaning representations from sample utterances of an understanding model to yield understanding model).
Sapugay also teaches performing a meaning search to extract an intent and entity of a received user utterance using the understanding model (Sapugay, para 0023, the meaning search subsystem compares meaning representations of the utterance meaning model to meaning representations of the understanding model to extract artifacts (e.g., intents and entities) from the received user utterance).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because the meaning extraction subsystem generates meaning representations from a received user utterance to yield an utterance meaning model and generates meaning representations from sample utterances of an understanding model to yield understanding model, and wherein the meaning search subsystem compares meaning representations of the utterance meaning model (Sapugay, para 0023).
Regarding Claim 17, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the method all the limitations of claim 13 upon which claim 17 depends, wherein the NLU framework comprises an intent-entity model, and the method comprises:
Sapugay, in the same field of endeavor, further teaches before performing the lookup source inference of the utterance, receiving the utterance from a user as a newly submitted sample utterance to be validated for inclusion in the intent-entity model (Sapugay, para 0065, the NLU framework processes the utterance based on the intent-entity model), and the at least one processor is configured to execute the stored instructions to cause the NLU framework to perform actions (Sapugay, para 0056, the one or more processors may include one or more microprocessors capable of performing instructions stored in the memory).
Sapugay also teaches after performing the vocabulary injection to generate the plurality of re-expressions of the utterance, providing, to the user (Sapugay, para 0094, block 308, performing vocabulary injection on the original utterance; Vocabulary injection generally involves introducing multiple re-renderings of the original utterance; para 0098, an utterance generated from vocabulary injection), an indication that the newly submitted sample utterance is valid or invalid based on the plurality of taxonomy segmentations (Sapugay, para 0078, Figure 7, the prosody subsystem of the meaning extraction subsystem analyzes the prosody of the utterance using a combination of rule-based and ML-based prosody plug-ins. Using these plug-ins, the prosody subsystem analyzes the utterance for prosodic cues, including written prosodic cues such as rhythm (e.g., chat rhythm, such as utterance bursts, segmentations).
Sapugay also teaches providing, to the user, the plurality of re-expressions of the utterance as suggestions of alternative valid sample utterances for inclusion in the intent-entity model (Sapugay, para 0092, Figure 11; the NLU framework generates re-expressions of an original utterance, and then generates a set of meaning representations based on these re-expressions and the original utterance. The original utterance may be one of the sample utterances of the intent-entity model).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because the vocabulary subsystem may access the vocabulary model of the understanding model to identify alternative vocabulary that can be used to generate re-expressions of the utterances having different tokens (Sapugay, para 94).
Regarding Claim 19, Galitsky discloses a non-transitory, computer-readable medium storing instructions executable by a processor of a natural language understanding (NLU) framework, the instructions comprising instructions to:
Galitsky further discloses to compile a respective source data representation of a plurality of taxonomy lookup sources (Galitsky, para 263, database lookups) from taxonomy source data (Galitsky, para 228, taxonomy-based);
Galitsky does not specifically disclose performing vocabulary injection to generate a plurality of re-expressions of the utterance based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens indicated by a particular taxonomy segmentation with the corresponding alternative tokens indicated by the particular taxonomy segmentation.
However, Sapugay, in the same field of endeavor, teaches performing vocabulary injection (Sapugay, para 0094, performing vocabulary injection) to generate a plurality of re-expressions of the utterance (Sapugay, para 0094, generate re-expressions of the utterances) based on the plurality of taxonomy segmentations, wherein each of the plurality of re-expressions of the utterance substitute at least one of the matched tokens ( Sapugay, para 0094, generate a new utterance in which the term “developer” is substituted by the term “employee”) indicated by a particular taxonomy segmentation with the corresponding alternative tokens (Sapugay, para 0093; involve substituting certain tokens with other tokens) indicated by the particular taxonomy segmentation (Sapugay, para 0078; segmentations).
Galitsky in view of Sapugay do not specifically disclose source data that represents one or more hierarchal relationships between product categories within a domain of a client, wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories.
However, Keysar, in the same field of endeavor, discloses source data that represents one or more hierarchal relationships between product categories within a domain of a client (Keysar, –para 0023, Fig. 2, Fig. 2 shows an illustrative example of answering a question in accordance with some implementations of the present disclosure. In the illustrated example, a webpage search engine receives a “who” question, retrieves a number of search results, and provides an answer based on the search results and information associated with the search results. It will be understood that while the illustrated example refers to person entities in response to a “who” question, any suitable type of entity may be identified in response to any suitable type of question. For example, location entities may be identified in response to a “where” question; Keysar, 0032, In some implementations, the system calculates the sum the entity references occurring in list 230, list 232, and list 234, as shown in summation 236. In the illustrated example, there are 7 instances of [Juan Carlos I] identified in the top three search results shown in search result box 206. In some implementations, the system identifies [Juan Carlos I] as entity result 240 and provides answer 242 based on entity result 240. In some implementations, the system displays answer 242 including the text [The King of Spain is Juan Carlos I]. In some implementations, the system displays only text [Juan Carlos I]. In some implementations, the system generates a natural language or other format response based in part on the received query. In some implementations, the system may display additional information associated with the answer, for example, a picture of Juan Carlos I or a link to an encyclopedia entry. It will be understood that in some implementations, the system may display content in search results box 206 before displaying answer 240; Keysar, para 0049 - 0052, FIG. 4 shows illustrative knowledge graph 400 containing nodes and edges. Illustrative knowledge graph 400 includes nodes 402, 404, 406, and 408. Knowledge graph 400 includes edge 410 connecting node 402 and node 404. Knowledge graph 400 includes edge 412 connecting node 402 and node 406. Knowledge graph 400 includes edge 414 connecting node 404 and node 408. Knowledge graph 400 includes edge 416 and edge 418 connecting node 402 and node 408. Knowledge graph 400 includes edge 420 connecting node 408 to itself. Each aforementioned group of an edge and one or two distinct nodes may be referred to as a triple or 3-tuple. As illustrated, node 402 is directly connected by edges to three other nodes, while nodes 404 and 408 are directly connected by edges to two other nodes. Node 406 is connected by an edge to only one other node, and in some implementations, node 406 is referred to as a terminal node. As illustrated, nodes 402 and 408 are connected by two edges, indicating that the relationship between the nodes is defined by more than one property. As illustrated, node 408 is connected by edge 420 to itself, indicating that a node may relate to itself. While illustrative knowledge graph 400 contains edges that are not labeled as directional, it will be understood that each edge may be unidirectional or bidirectional. It will be understood that this example of a graph is merely an example and that any suitable size or arrangement of nodes and edges may be employed; Generally, nodes in a knowledge graph can be grouped into several categories. Nodes may represent entities, organizational data such as entity types and properties, literal values, and models of relationships between other nodes. A node of a knowledge graph may represent an entity, as defined above. In some implementations, entity types, properties, and other suitable content is created, defined, redefined, altered, or otherwise generated by any suitable technique. For example, content may be generated by manual user input, by automatic responses to user interactions, by importation of data from external sources, by any other suitable technique, or any combination thereof. For example, if a commonly searched-for term is not represented in the knowledge graph, one or more nodes representing that node may be added. In another example, a user may manually add information and organizational structures. A node representing organizational data may be included in a knowledge graph. These may be referred to herein as entity type nodes).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Keysar in the method of Galitsky in view of Sapugay because that would enable to rank the entity references and also to select any entity reference(s) from the ranking list in the determination of an answer (Keysar, Abstract).
Galitsky in view of Sapugay and Keysar do not specifically disclose wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories.
However, Yamada, in the same field of endeavor discloses wherein the hierarchal relationships indicate hypernyms, hyponyms, or both of the product categories (Yamada, p. 934, 1st col, Figure 4: [i.e., hierarchal relationships indicating the hypernym car and the hyponyms hybrid vehicle, mini vehicle, etc., which are also product categories]).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Yamada in the method of Galitsky in view of Sapugay and Keysar because that would extract a hierarchical structure relations that produces a hypernym of a certain word that may also be the hyponym of another hypernym with a deeper hierarchical structure method (Yamada, p. 931, 2nd col, 2nd para).
Galitsky in view of Sapugay, Keysar and Yamada do not specifically disclose to perform lookup source inference of different portions of an utterance across the plurality of taxonomy lookup sources using parallel processing threads to determine a plurality of taxonomy segmentations for the utterance, wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources, and indicates corresponding alternative tokens for each of the matched tokens of the utterance.
However, Lunenfeld, in the same field of endeavor, discloses to perform lookup source inference of different portions of an utterance across the plurality of taxonomy lookup sources using parallel processing threads to determine a plurality of taxonomy segmentations for the utterance (Lunenfeld, Para 0826 – 0828, parallel processing of multiple queries/keyword searches of multiple information sources of the same and/or different types and may be used on substantially any kind of network; quick response intelligence gathering of multiple same and/or different information requests of multiple sources, grouping and sorting results substantially simultaneously in real time and on-the-fly; combined search and E-Commerce, and/or as a single point of purchase/sale for multiple products in multiple categories from multiple sites, and is particularly useful for corporate, industrial, commercial, and government purchasing of multiple products from multiple sources, as well as internet purchasing of multiple products from multiple sources ), wherein each of the plurality of taxonomy segmentations indicates how tokens of the utterance match to the respective source data representation of one or more of the plurality of taxonomy lookup sources, and indicates corresponding alternative tokens for each of the matched tokens of the utterance (Lunenfeld, Para 0829-0831, performing research, using multiple information sources, multiple sites, search engines, servers, databases, clients, applications, software applications, programs, and/or software programs, and may be performed in parallel using multiple queries/keyword phrases in multiple categories and/or multiple fields substantially simultaneously, in real time, and on-the-fly; downloading multiple title/subject and/or music/audio/video/television substantially simultaneously; presenting results to single and/or multiple users substantially simultaneously in real time and on-the-fly ).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Lunenfeld in the method of Galitsky in view of Sapugay, Keysar and Yamada because that would enable retrieving substantially multiple simultaneous services and/or information having the same and/or different criteria from the same and/or different servers, sorting, grouping, and/or organizing the responses from the servers and/or the clients into information and/or services responses, and communicating the service and/or information responses to the requesters and/or the users substantially simultaneously (Lunenfeld, para 0008).
Regarding Claim 20, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the medium of all the limitations of claim 19 upon which claim 20 depends, wherein the utterance is a received user utterance (Galitsky, para 116, user saying).
Sapugay further discloses a sample utterance of an intent-entity model of the NLU framework (Sapugay, para 0013, NLU framework that includes a meaning extraction subsystem capable of generating multiple meaning representations for utterances, including sample utterances in the intent-entity model and utterances received from a user), or a newly submitted sample utterance to be validated for inclusion in the intent-entity model (Sapugay, para 0092, Figure 11; the NLU framework generates re-expressions of an original utterance, and then generates a set of meaning representations based on these re-expressions and the original utterance. The original utterance may be one of the sample utterances of the intent-entity model).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Sapugay in the method of Galitsky because this will enable the agent automation framework that includes an NLU framework and an intent-entity model having defined intents and entities that are associated with sample utterances (Sapugay, para 0010).
Regarding Claim 21, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the NLU framework of claim 1.
Furthermore, Yamada discloses:
wherein the hypernyms, hyponyms, or both of the product categories are specific to the domain of the client (Yamada, p. 934, 1st col, Figure 4: [i.e., hierarchal relationships indicating the hypernym car and the hyponyms hybrid vehicle, mini vehicle, etc., which are also product categories]; Yamada, p. 932, 1st col, 1st para, The distributional hypothesis states that words that occur in similar contexts tend to be semantically similar (Harris 1985). In this section, we first introduce distributional similarity based on raw verb-noun dependencies (RVD). To avoid the sparseness problem of the co-occurrence of verb-noun dependencies, we also use distributional similarity based on a large-scale clustering of verb-noun dependencies (CVD); Yamada, page 929, 2nd col, 3rd para, In the experiment, we extracted hypernyms for approximately 670,000 target words that are not included in the Wikipedia relation database but are found on the Web. We tested two distributional similarities: one based on raw verb-noun dependencies (RVD) and the other based on a large-scale clustering of verb-noun dependencies (CVD); Yamada, page 933, 1st col, 4th para - page 934, 1st col, 1st para, In the Wikipedia relation database, there are about 95,000 hypernyms and about 1.2 million hyponyms. In both RVD and CVD, the words used were selected according to the number (the number of kinds, not the frequency) of <v, rel >s that n has dependencies in the data. As a result, 1 million words were selected. The number of common words that are also included in the Wikipedia relation database are as follows: Hypernyms 28,015 (common hypernyms) Hyponyms 175,022 (common hyponyms) These common hypernyms become candidates for hypernyms for a target word. On the other hand, the common hyponyms are used as clues for identifying appropriate hypernyms… As a result of scoring, each hypernym has a score for the target word. The hypernym that has the highest score for the target word is selected as its hypernym. The hyponymy relations thus produced are ranked according to the scores. Figure 4 shows an example of the scoring process. In this example, we use CitroenAX as the target word whose hypernym will be identified. First, the k similar words are extracted from the common hyponyms in the Wikipedia relation: Opel Astra, TVR Tuscan, Mitsubishi Minica, and Renault Lutecia are extracted. Next, each k similar word votes a score to its ancestors. The words Opel Astra, TVR Tuscan, and Renault Lutecia vote to their parent car and the word Mitsubishi Minica votes to its parent mini-vehicle and its grandparent car with a small penalty. Finally, the hypernym car, which has the highest score, is selected as the hypernym of the target word CitroenAX; [i.e., the Wikipedia relation database can be the client, and as in Figure 4 CitroenAX is the target word whose hypernym is to be identified and CitroenAX for “car” product categories is specific product name]).
Regarding Claim 22, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the NLU framework of claim 21.
Furthermore, Sapugay, discloses:
wherein the taxonomy segmentations, the alternative tokens, or both are specific to the domain of the client (Sapugay, para 0092-0094, generate re-expressions of the utterances; Sapugay, para 0094, generate a new utterance in which the term “developer” is substituted by the term “employee”; Sapugay, para 0093, For the embodiment illustrated in FIG. 11, the process 300 begins with the vocabulary subsystem 170 of the NLU framework 104 cleansing (block 306) the original utterance 302. For example, the vocabulary subsystem 170 may access and apply rules 114 stored in the database 106 to modify certain tokens (e.g., words, phrases, punctuation, emojis) of the utterance. For example, in certain embodiments, cleansing may involve applying a rule that removes non-textual elements (e.g., emoticons, emojis, punctuation) from the original utterance 302. In certain embodiments, cleansing may involve correcting misspellings or typographical errors in the utterance. Additionally, in certain embodiments, cleansing may involve substituting certain tokens with other tokens. For example, the vocabulary subsystem 170 may apply a rule that that all entities with references to time or color with a generic or global entity (e.g., “TIME”, “COLOR”); [i.e., in Figure 11, modify or replace “certain tokens (e.g., words, phrases, punctuation, emojis) of the utterance” according set rules of “the vocabulary subsystem 170 of the NLU framework 104 cleansing (block 306) the original utterance 302.” As an example, the system ‘generates a new utterance in which the term “developer” is substituted by the term “employee” ‘; “the vocabulary subsystem” as the “domain”]).
Regarding Claim 23, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the NLU framework of claim 1.
Furthermore, Yamada, discloses:
wherein the hierarchal relationships indicate synonyms, formal names, or colloquial names of the product categories that are specific to the domain of the client (Yamada, p. 934, 1st col, Figure 4: [i.e., hierarchal relationships indicating the hypernym car and the hyponyms hybrid vehicle, mini vehicle, etc., which are also product categories]; In the Wikipedia relation database, there are about 95,000 hypernyms and about 1.2 million hyponyms. In both RVD and CVD, the words used were selected according to the number (the number of kinds, not the frequency) of <v, rel >s that n has dependencies in the data. As a result, 1 million words were selected. The number of common words that are also included in the Wikipedia relation database are as follows: Hypernyms 28,015 (common hypernyms) Hyponyms 175,022 (common hyponyms) These common hypernyms become candidates for hypernyms for a target word. On the other hand, the common hyponyms are used as clues for identifying appropriate hypernyms… As a result of scoring, each hypernym has a score for the target word. The hypernym that has the highest score for the target word is selected as its hypernym. The hyponymy relations thus produced are ranked according to the scores. Figure 4 shows an example of the scoring process. In this example, we use CitroenAX as the target word whose hypernym will be identified. First, the k similar words are extracted from the common hyponyms in the Wikipedia relation: Opel Astra, TVR Tuscan, Mitsubishi Minica, and Renault Lutecia are extracted. Next, each k similar word votes a score to its ancestors. The words Opel Astra, TVR Tuscan, and Renault Lutecia vote to their parent car and the word Mitsubishi Minica votes to its parent mini-vehicle and its grandparent car with a small penalty. Finally, the hypernym car, which has the highest score, is selected as the hypernym of the target word CitroenAX; [i.e., the target word can be the client and in Figure 4, CitroenAX is considered the target word whose hypernym is to be identified]).
Regarding Claim 24, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the NLU framework of claim 1.
Furthermore, Lunenfeld teaches
wherein the parallel processing threads comprise a plurality of separate processing threads implemented in parallel with one another to simultaneously perform the lookup source inference on the different portions of the utterance across the plurality of taxonomy lookup sources (Lunenfeld, Para 0826 – 0831, parallel processing of multiple queries/keyword searches of multiple information sources of the same and/or different types and may be used on substantially any kind of network; quick response intelligence gathering of multiple same and/or different information requests of multiple sources, grouping and sorting results substantially simultaneously in real time and on-the-fly; combined search and E-Commerce, and/or as a single point of purchase/sale for multiple products in multiple categories from multiple sites, and is particularly useful for corporate, industrial, commercial, and government purchasing of multiple products from multiple sources, as well as internet purchasing of multiple products from multiple sources; performing research, using multiple information sources, multiple sites, search engines, servers, databases, clients, applications, software applications, programs, and/or software programs, and may be performed in parallel using multiple queries/keyword phrases in multiple categories and/or multiple fields substantially simultaneously, in real time, and on-the-fly; downloading multiple title/subject and/or music/audio/video/television substantially simultaneously; presenting results to single and/or multiple users substantially simultaneously in real time and on-the-fly ).
Claims 10, 12, 14 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Galitsky in view of Sapugay, further in view of Keysar, further in view of Yamada, further in view of Lunenfeld as applied to claims 1 and 13 above, and further in view of Hancock Pub. No. US 2020/0050638 A1 (herein after called Hancock).
Regarding Claim 10, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of claim 1 upon which claim 10 depends.
wherein to select the corresponding alternative token, the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions (Galitsky para 263, database lookups; Galitsky para 539, system memory may store program instructions that are loadable and executable on processing unit (i.e., at least one processor).
Sapugay further teaches to perform vocabulary injection (Sapugay, para 0094, performing vocabulary injection), the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions (Sapugay, para 0056, one or more microprocessors capable of performing instructions stored in the memory) comprising:
Sapugay further also further teaches for each taxonomy segmentation of the plurality of taxonomy segmentations of the utterance (Sapugay, para 0078, analyzes the utterance for prosodic cues, including written prosodic cues such as rhythm (e.g., chat rhythm, such as utterance bursts, segmentations)).
Sapugay further teaches generating one of the plurality of re-expressions of the utterance that substitutes the matched token of the utterance with the corresponding alternative token (Sapugay, para 94, generate re-expressions of the utterances having different tokens).
Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld do not specifically teach for each matched token of the matched tokens of the utterance indicated by the taxonomy segmentation. Galitsky in view of Sapugay Keysar do not also specifically teach selecting, based on a configuration of the NLU framework, a corresponding alternative token from the corresponding alternative tokens for the matched token indicated by the taxonomy segmentation, and in response to determining that the corresponding alternative token complies with the level of hypernym-based replacement, selecting the corresponding alternative token from the corresponding alternative tokens indicated by the taxonomy segmentation.
However, Hancock, in the same field of endeavor, teaches:
for each matched token of the matched tokens of the utterance indicated by the taxonomy segmentation. Hancock also teaches selecting, based on a configuration of the NLU framework, a corresponding alternative token from the corresponding alternative tokens for the matched token indicated by the taxonomy segmentation, wherein the configuration defines a level of hypernym-based replacement (Hancock, para 0081; employ a taxonomy-based keyword extraction method. In such a method, a segment of text is analyzed to determine whether it matches one or more of a predefined set of keywords (i.e., tokens); para 0113, The tokenizer module can tokenize the text data produced by the text preprocessor into lexical units, such as paragraphs, sentences, or words; para 0126, The tokenizers can take text fields of search documents and divide them into one or more “tokens,” where each token comprises a lexical unit—e.g., a word, sentence, phrase, or paragraph).
in response to determining that the corresponding alternative token complies with the level of hypernym-based replacement, selecting the corresponding alternative token from the corresponding alternative tokens indicated by the taxonomy segmentation (Hancock, para 0124, TABLE-US-00006, match any token sequence on the LHS of “=>” | # and replace with all alternatives on the RHS; Hancock, para 0103, synonyms can be obtained by a synonym module by obtaining the list of synonyms, holonyms, hypernyms; i.e., hypernym-based replacement by a synonym; Hancock, para 0081; employ a taxonomy-based keyword extraction method. In such a method, a segment (i.e., segmentation) of text is analyzed to determine whether it matches one or more of a predefined set of keywords (i.e., tokens); para 0126, The tokenizers can take text fields of search documents and divide them into one or more “tokens,” where each token comprises a lexical unit—e.g., a word, sentence, phrase, or paragraph; Hancock, para 0103, synonyms can be generated using a lexical database, containing a list of words, with data about the words. The data can comprise one or more of a definition, part of speech, list of synonyms, similar words, hypernym; Hancock, para 0113, In some embodiments, the natural language processing module 603 can process textual matter and/or metadata to produce a portion of the query 602. The natural language processing module 603 can comprise a text preprocessor module 605, a tokenizer module 606, a synonym generator module 607, a machine translation module 608, a stop word editor 609, and a key phrase extractor 610. The text preprocessor module 605 can receive text data and perform preprocessing steps, such as encoding conversion (e.g. ASCII to Unicode), whitespace removal/replacement, removal of invalid or irrelevant characters, and other text data normalization steps. The tokenizer module 606 can tokenize the text data produced by the text preprocessor 605 into lexical units, such as paragraphs, sentences, or words. The synonym generator 607 can analyze textual data and produce a list of relevant synonyms, such as a list of synonyms for words occurring in the textual data in accordance with the techniques for generating synonyms discussed infra, based on the contents of the query data and/or other sources of text, such as standard corpora for a language or corpora derived from a set of patents in the same class, subclass, or related or similar classes, all patent documents, or a set of documents in a similar technical field (such as journal articles, textbooks, product documentation, etc.); [i.e., “corresponding alternative token complies with the level of hypernym-based replacement” ]).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Hancock in the method of Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld because this will enable a citation generator to generate human-readable citations to tokens produced by the tokenizer (Hancock, para 0099).
Regarding Claim 12, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of claim 10 upon which claim 12 depends.
Galitsky in view of Sapugay, Keysar and Yamada do not specifically teach the configuration that indicates substitution of the matched token with the corresponding alternative tokens that are hypernyms, or synonyms, or formal names, or colloquial names of the matched token, and wherein, to select the corresponding alternative token, the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions, and in response to determining that the corresponding alternative token is a hypernym, or a synonym, or a formal name, or a colloquial name of the matched token, in accordance with the configuration, selecting the corresponding alternative token from the corresponding alternative tokens indicated by the taxonomy segmentation.
However, Hancock, in the same field of endeavor, teaches:
the configuration that indicates substitution of the matched token with the corresponding alternative tokens that are hypernyms, or synonyms, or formal names, or colloquial names of the matched token, and wherein, to select the corresponding alternative token, the at least one processor is configured to execute the stored instructions to cause the NLU system to perform actions (Hancock, para 0079; substitution of words or phrases with equivalent words or phrases; para 0081; a segment (i.e., segmentation) of text is analyzed to determine whether it matches one or more of a predefined set of keywords (i.e., tokens); para 0126, The tokenizers can take text fields of search documents and divide them into one or more “tokens,” where each token comprises a lexical unit—e.g. a word, sentence, phrase, or paragraph; Hancock, para 0103, synonyms can be generated using a lexical database, containing a list of words, with data about the words. The data can comprise one or more of a definition, part of speech, list of synonyms, similar words, hypernym).
the corresponding alternative token is a hypernym, or a synonym, or a formal name, or a colloquial name of the matched token, in accordance with the configuration, selecting the corresponding alternative token from the corresponding alternative tokens indicated by the taxonomy segmentation (Hancock, para 0124, match any token; para 0125, Synonym token filter; para 0127, Synonym filters can convert words based on a synonym list either by replacing words with a common synonym; para 0102, a synonym generator that generates a list of relevant synonyms; para 0103, synonyms can be obtained by a synonym module by obtaining the list of synonyms, holonyms, hypernym).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Hancock in the method of Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld because this will enable the synonyms to be generated by the synonym generator that can be output to the search engine for use in a synonym list module and/or synonym filter in filters (Hancock, para 0113).
Regarding Claim 14, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose all the limitations of the method of claim 13 upon which claim 14 depends.
Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld do not specifically teach the taxonomy source data (Hancock, para 0081, taxonomy-based) comprises a table having a plurality of vocabulary source columns and a hypernym column, and the method, grouping the plurality of vocabulary source columns of the taxonomy source data using the hypernyms column as a pivot column to generate hypernym-grouped taxonomy source data table, wherein each entry the hypernym-grouped taxonomy source data table comprises a unique hypernym token and a set of alternative tokens related to the unique hypernym token.
However, Hancock, in the same field of endeavor, teaches:
the taxonomy source data (Hancock, para 0081, taxonomy-based) comprises a table having a plurality of vocabulary source columns and a hypernym column, and the method (Hancock, para 0103, list of synonyms, holonyms, hypernyms; para 0133, words in the vocabulary).
grouping the plurality of vocabulary source columns of the taxonomy source data using the hypernyms column (Hancock, para 0081, taxonomy-based; para 0103, list of synonyms, holonyms, hypernyms; para 0133, words in the vocabulary) as a pivot column to generate hypernym-grouped taxonomy source data table, wherein each entry the hypernym-grouped taxonomy source data table comprises a unique hypernym token and a set of alternative tokens related to the unique hypernym token (Hancock, para 0126, The tokenizers can take text fields of search documents and divide them into one or more “tokens;” para 0081, taxonomy-based; para 0103, the data can comprise one or more of a definition, part of speech, list of synonyms, hypernym; para 0133, words in the vocabulary).
Hancock also teaches compiling the respective source data representation of each of the plurality of taxonomy lookup sources from a particular entry in the hypernym-grouped taxonomy source data table (Hancock, para 0101, generate a data representation that contains the constituent parts of the citation, such as page, paragraph, column, line or other numbering schemes. Storing a data representation; para 0061, the disclosed technology is disclosed in terms of modules and submodules, each of which are to be understood as discrete units of functionality, which can be embodied as classes, modules, functions, compilation; para 0103, list of synonyms, holonyms, hypernyms).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Hancock in the method of Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld because compiling a data representation can ease generation of final citations in output reports by allowing a group of contiguous search documents to be easily combined into a single citation (Hancock, para 0101).
Regarding Claim 18, Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld disclose the method of all the limitations of claim 13 upon which claim 18 depends. Sapugay further teaches performing vocabulary injection (Sapugay, para 0094, performing vocabulary injection) comprises:
Sapugay also further teaches for each taxonomy segmentation of the plurality of taxonomy segmentations (Sapugay, para 0078, analyzes the utterance for prosodic cues, including written prosodic cues such as rhythm (e.g., chat rhythm, such as utterance bursts, segmentations)).
Sapugay further teaches generating one of the plurality of re-expressions of the utterance that substitutes the matched token of the utterance with the corresponding alternative token (Sapugay, para 94, generate re-expressions of the utterances having different tokens).
Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld do not specifically teach for each matched token of the matched tokens of the utterance indicated by the taxonomy segmentation, and for each corresponding alternative token indicated by the taxonomy segmentation.
However, Hancock, in the same field of endeavor, teaches:
for each matched token of the matched tokens of the utterance indicated by the taxonomy segmentation (Hancock, para 0081; employ a taxonomy-based keyword extraction method. In such a method, a segment of text is analyzed to determine whether it matches one or more of a predefined set of keywords (i.e., tokens); para 0113, The tokenizer module can tokenize the text data produced by the text preprocessor into lexical units, such as paragraphs, sentences, or words; para 0126, The tokenizers can take text fields of search documents and divide them into one or more “tokens,” where each token comprises a lexical unit—e.g., a word, sentence, phrase, or paragraph)..
for each corresponding alternative token indicated by the taxonomy segmentation (Hancock, para 0081; employ a taxonomy-based keyword extraction method. In such a method, a segment of text is analyzed to determine whether it matches one or more of a predefined set of keywords (i.e., tokens); para 0113, The tokenizer module can tokenize the text data produced by the text preprocessor into lexical units, such as paragraphs, sentences, or words; para 0126, The tokenizers can take text fields of search documents and divide them into one or more “tokens,” where each token comprises a lexical unit—e.g., a word, sentence, phrase, or paragraph).
Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate the method of Hancock in the method of Galitsky in view of Sapugay, Keysar, Yamada and Lunenfeld because this will enable a citation generator to generate human-readable citations to tokens produced by the tokenizer (Hancock, para 0099).
Allowable Subject Matter
Claim 3 is objected to as being dependent upon rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims, and also if all these claims overcome the 101 rejections. The reasons for allowance are that the prior art of record do not specifically teach the limitations as recited in claim 3.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Shank (US 20210383205 A1) discloses taxonomy Construction Via Graph-Based Cross-domain Knowledge Transfer. Here a system, computer program product, and method are provided for employing a graph neural network (GNN) to construct a taxonomy. The GNN is subject to a training cycle and an inference cycle. The training cycle encodes cross-domain terms pairs from a set of noisy cross domain pairs extracted from a corpora, and outputs a preliminary taxonomy. The inference cycle identifies candidate term pairs and selectively subjects the candidate term pairs to selective filtering to produce a system predicted taxonomy from the preliminary taxonomy.
Krishnamurthy (US 10496749 B2) teaches Unified Semantics-focused Language Processing and Zero Base Knowledge Building System. Here a method and a language processing and knowledge building system (LPKBS) for processing textual data, receives textual data and a language object; segments the textual data into sentences and each sentence into words; generates a list of one or more natural language phrase objects (NLPOs) for each word by identifying vocabulary classes and vocabulary class features for each word based on vocabulary class feature differentiators; creates sentence phrase lists, each including a combination of one NLPO selected per word from each list of NLPOs.
Galitsky (US 11875118 B2) DISCLOSES Detection of Deception Within Text Using Communicative Discourse. Systems, devices, and methods of the present invention detect deceptive or fake content in text. In an example, a computer system generates, from text a discourse tree that represents rhetorical relationships between fragments of the text. The computer system generates a communicative discourse tree from the discourse tree. The computer system identifies a number of non-trivial rhetorical relations associated with the nonterminal nodes in the communicated discourse tree and, for each terminal edge having a communicative action, a level of nesting of the communicative action. The computer system derives, from the number of non-trivial rhetorical relations and the levels of nesting of the identified communicative actions, a complexity score that is indicative of a level of deception in the text.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MULUGETA T. DUGDA whose telephone number is (703)756-1106. The examiner can normally be reached Mon - Fri, 4:30am - 7:00pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Paras D. Shah can be reached on 571-270-1650. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MULUGETA TUJI DUGDA/Examiner, Art Unit 2653
/Paras D Shah/Supervisory Patent Examiner, Art Unit 2653
03/04/2026