DETAILED ACTION
The present office action is responsive to the applicant’s filing on 01/02/2025.
The application has claims 1-10 present. All present claims have been examined.
The IDS document and cited references filed 01/02/2025 have been reviewed and considered by the examiner.
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 .
Examiner Notes
Examiner cites particular columns, paragraphs, figures and line numbers in the references as applied to the claims below for the convenience of the applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply as well. It is respectfully requested that, in preparing responses, the applicant fully consider the references in their entirety as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art or disclosed by the examiner. The entire reference is considered to provide disclosure relating to the claimed invention. The claims & only the claims form the metes & bounds of the invention. Office personnel are to give the claims their broadest reasonable interpretation in light of the supporting disclosure. Unclaimed limitations appearing in the specification are not read into the claim. Prior art was referenced using terminology familiar to one of ordinary skill in the art. Such an approach is broad in concept and can be either explicit or implicit in meaning. Examiner's Notes are provided with the cited references to assist the applicant to better understand how the examiner interprets the applied prior art. Such comments are entirely consistent with the intent & spirit of compact prosecution.
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 1-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
Claim 1-7 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter because the claimed invention is directed to a judicial exception (i.e. an abstract idea) without significantly more.
Claim 1 recites a “Method” comprising steps that may be mental process (prong 1). The overall process presented in the claim is to summarize text from a document. The text part of the document is summarized based on classification, ordering and clustering of the related sentences using context and semantical similarities. See par. 5-6 of the specification. The steps taken provide an opportunity to select appropriate sentences that provide a summary without losing important context.
Limitations under Step 2A - prong 1:
The specific limitations of
“extracting, by the text summarizing device, one or more keywords..”; (is a mental process; a person can mentally extract text from a document or do this via an aid).
“identifying, by the text summarizing device…, a set of additional keywords..” (is a mental process; determine other additional words associated to first keywords).
“semantically scoring, by the text summarizing device, sentences…(is a mental process; assigning values to sentences based on certain criteria e.g. similarity).
“ranking, by the text summarizing device, each of the plurality of sentences…” (is a mental process; organizing/importance for the order of the sentences based on previous determination or a desired criteria).
“performing, by the text summarizing device, a contextual classification…(is a mental process; further organizing the sentences based on criteria).
“clustering, by the text summarizing device, each of the set of sentences… (is a mental process; further organizing the sentences based on criteria to make sense of the text).
“generating, by the text summarization device, a consolidated summarized text…” (is a mental process with an aid; a person can mentally organize the text to write the summary; the limitation can be simply be writing it down via an aid)
Limitations under Step 2A - prong 2:
The limitations of
The limitations of "receiving…. a text article from a document”, is insignificant extra-solution activity to the judicial exception, as mere data gathering (See MPEP 2106.05(g)).
The limitations to include “distance calculation”, “using a similarity algorithm”, “vectorization,” and word embedding” is interpreted as merely using instructions or a computer as a tool to perform the abstract idea (see MPEP 2106.05 (f)).
Step 2B – not significant more.
Thus, the recited “Method” is an abstract idea in that it is not tied to a particular machine or apparatus and it does not transform a particular article into a different state or thing. Furthermore, the additional element of using a computer as a tool to perform the recited steps amounts to no more than mere instructions to apply the abstract idea using a generic computer component. Mere instructions to apply a judicial exception using a generic computer component cannot provide an inventive concept.
Accordingly, the recited method is non-statutory subject matter.
Claim 2: wherein the set of additional keywords comprises a plurality of affinity keywords, a plurality of semantically significant keywords, and a plurality of similar keywords, and wherein the set of additional keywords provide additional information that is not captured by the one or more keywords (a person can look at the content and determine similarities or associations). As such, it further describes the abstract idea previously identified in the independent claims. Thus, the claims recite an abstract idea and are not patent-eligible.
Claim 3: wherein the semantically scoring comprises: assigning a weight to each of the set of additional keywords; and calculating a semantic score for each sentence based on the weight assigned to each of the set of additional keywords. (a person can assign values and determine similarities and importance to desired text that wishes to summarize). As such, it further describes the abstract idea previously identified in the independent claims. Thus, the claims recite an abstract idea and are not patent-eligible.
Claim 4: wherein: the contextual classification is one of a rule-based classification or a model- based classification…(is interpreted as simply having a machine do a classification. This would fall under “apply it” or having mere instructions implement the abstract idea). Thus, the claims recite an abstract idea and are not patent-eligible.
Claim 5: wherein clustering the set of sentences based on the semantic scoring comprises: identifying one or more sentences within the set of sentences having relevant semantic scores; and grouping each of the one or more sentences into a separate cluster based on the identifying, wherein each of the separate cluster comprises the one or more sentences with a corresponding relevant semantic score (further explain providing values, grouping based on criteria like importance or relevance) Thus, the claims recite an abstract idea and are not patent-eligible.
Claim 6: wherein clustering the set of sentences based on the embedding comprises: identifying one or more sentences within the set of sentences having relevant embeddings; and grouping each of the one or more sentences into a separate cluster based on the identifying, wherein each of the separate cluster comprises the one or more sentences with a corresponding relevant embeddings (is interpreted as simply having a machine do the steps for grouping based on certain criteria. This would fall under “apply it” or having mere instructions implement the abstract idea). Thus, the claims recite an abstract idea and are not patent-eligible.
Claim 7: wherein generating the consolidated summarized text comprises: generating a summarized text for each of the separate cluster; and concatenating the summarized text of each of the separate cluster to obtain a consolidated summarized text (further describes the abstract idea previously identified in the independent claims. Thus, the claims recite an abstract idea, explain a mental process with an aid and are not patent-eligible. As a person can mentally organize the text and decide what to write in the summary; the limitation can be simply be writing it down via an aid)
Claim 8-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter because the claimed invention is directed to a judicial exception (i.e. an abstract idea) without significantly more.
Claims 8-10 provide the limitations in a similar manner as those provided from claims 1-3 in a “system” embodiment. As such these claims are interpreted along the same rational as claims 1-3 above.
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.
Claim(s) 1, 2, 8 and 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kandekar et al. (US 20120210203), in view of Chhabra et al. (US 20200293618).
In regards to claims (1 and 8), Kandekar a method for summarizing text articles of documents, the method comprising: receiving, by a text summarizing device, a text article from a document, wherein the text article comprises a plurality of sentences (see FIG. 2 and at least para 35-36, 40: summarizing text with sentences e.g. ebook “War and Peace”); extracting, by the text summarizing device, one or more keywords from the plurality of sentences based on a keyword extraction algorithm (extracting keywords. See at least para 35-37, 42: “extraction-based methods are used for summary generation, whereby content segments, such as passages, sentences, phrases or words”); semantically scoring, by the text summarizing device, the plurality of sentences based on a weight of each set of the keywords and a frequency of each word in the plurality of sentences (semantic analysis and weights to further select sentences. See at least para 42, 44, 48, 76, 95: “key phrases, words, and/or sentences in content may be identified and used in a summary for the content. The length of the content summary may be adapted by varying constraints or thresholds on the ranks or scores of phrases, words, and/or sentences in the content. Low ranking or scoring phrases, words, and/or sentences may be extracted, while others are used in the summary. Phrases, words, and/or sentences may be extracted from use in the summary based on a relevancy level assigned to the phrases, words, and/or sentences. As previously described, ranks or weights of the extracted phrases, words or sentences may also depend on the user profile”. Also, ”the media summary generator 118 may apply a semantic analysis technique to a portion of the e-book 116 for generating a content summary)”; ranking, by the text summarizing device, each of the plurality of sentences based on the semantically scoring; performing, by the text summarizing device, a contextual classification to select a set of sentences from the plurality of sentences based on one of the semantically scoring or the ranking (ranking sentences to group and select to generate summary. See at least para 42, 63-67, 83: “includes ranking the identified summary segments of step 510. For example, if more than one summary segment was identified for the content segment, the media summary generator 118 may rank the summary segments based on one or more predefined criteria”); clustering, by the text summarizing device, each of the set of sentences based on one of the semantic scoring or embedding to generate a summarized text for each cluster; and generating, by the text summarizing device, a consolidated summarized text based on the clustering (Clustering based on semantic analysis and generating summery from the clustered content. See para 70, 76, 81: “cluster (step 606) and applying an automated summarization technique to the cluster (step 608). For example, the media summary generator 118 may access a first cluster for the e-book 116, which corresponds to chapters 1-5. A summarization technique may be applied to the content of chapters 1-5. For example, the media summary generator 118 may apply a semantic analysis technique, an NLP technique, a summary extraction technique, the like, or combinations thereof to the content of chapters 1-5. The resulting summarization of the content associated with the cluster may be referred to as a "summary segment”. On para 81-82: “Steps 606-618 may be performed iteratively until there are no more clusters. In response to determining that there are no more clusters at step 618, the method may proceed to step 620. At step 620 of FIG. 6, the method includes presenting the content summary").
Kandekar doesn’t specifically teach identifying, by the text summarizing device and from each word of the plurality of sentences, a set of additional keywords corresponding to the one or more keywords based on one of a distance calculation, a similarity algorithm, a vectorization, or a word embedding techniques.
Chhabra teaches identifying, by the text summarizing device and from each word of the plurality of sentences, a set of additional keywords corresponding to the one or more keywords based on one of a distance calculation, a similarity algorithm, a vectorization, or a word embedding techniques (see at least para 32: teaches identifying words associated to determined keywords and using machine learning techniques which identify similarity between keywords to make selection used for generating the summary. “For instance, if there are several different entries, e.g., segments, that each state, “there is a problem with shipping,” “shipping is going to be delayed,” “we might have delays based on packaging,” and “shipping will not be on time, it will be late,” the number of occurrences of a particular word can be used to determine a priority for a keyword, and the priority can be compared against a threshold, e.g., a priority threshold. If the number of occurrences of a particular keyword exceeds the threshold, the system can determine that particular keyword as a topic… Other words, also referred to herein as supporting words, that surround the selected word can be used to generate sentences, such as “delay,” “late,” etc. … a threshold requirement can define any criteria that can be used to identify a relationship between any subset of segments or portions of a subset of segments. For instance, a subset of segments may meet a threshold requirement based on a number of phrases or keywords that align with a template or other keywords defined in one or more preferences. Any suitable machine learning techniques for identifying similarities between keywords and phrases may also be utilized”).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention these teachings of Chhabra in combination with Kandekar in order to identify additional keywords associated the sentence and keywords on those sentences, since it enhances the method for generating the summary by associating additional important words that can provide additional or a more complete context when selecting the portions for generating the summary.
In regards to claims (2 and 9), Kandekar doesn’t specifically teach wherein the set of additional keywords comprises a plurality of affinity keywords, a plurality of semantically significant keywords, and a plurality of similar keywords, and wherein the set of additional keywords provide additional information that is not captured by the one or more keywords.
Chhabra teaches wherein the set of additional keywords comprises a plurality of affinity keywords, a plurality of semantically significant keywords, and a plurality of similar keywords, and wherein the set of additional keywords provide additional information that is not captured by the one or more keywords (teaches identifying words, semantically significant words and similar words associated and using machine learning techniques which identify similarity between keywords (e.g. other determined keywords which are associated to a determined topic keyword), to make selection used for generating the summary. See at least para 32-33, 39: “a number of selected segments may be generated based on a priority of a particular topic. For instance, if there are several different entries, e.g., segments, that each state, “there is a problem with shipping,” “shipping is going to be delayed,” “we might have delays based on packaging,” and “shipping will not be on time, it will be late,” the number of occurrences of a particular word can be used to determine a priority for a keyword, and the priority can be compared against a threshold, e.g., a priority threshold. If the number of occurrences of a particular keyword exceeds the threshold, the system can determine that particular keyword as a topic… Other words, also referred to herein as supporting words, that surround the selected word can be used to generate sentences, such as “delay,” “late,” etc. … a threshold requirement can define any criteria that can be used to identify a relationship between any subset of segments or portions of a subset of segments. For instance, a subset of segments may meet a threshold requirement based on a number of phrases or keywords that align with a template or other keywords defined in one or more preferences. Any suitable machine learning techniques for identifying similarities between keywords and phrases may also be utilized”. Also on para 39: “For instance, if the selected sections indicate a particular topic having a threshold priority, any document content having keywords that are related to the topic can be extracted from the document and provided in the summary.”).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention these teachings of Chhabra in combination with Kandekar in order to identify additional keywords associated the sentence and keywords on those sentences, since it enhances the method for generating the summary by associating additional important words that can provide additional or a more complete context (e.g. related to a determined topic keyword), when selecting the portions for generating the summary.
Claim(s) 3 and 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kandekar et al. (US 20120210203), in view of Chhabra et al. (US 20200293618) and further in view of Malladi et al. (US 11741143).
In regards to claims 3 and 10, Kandekar doesn’t specifically teach wherein the semantically scoring comprises: assigning a weight to each of the set of additional keywords; and calculating a semantic score for each sentence based on the weight assigned to each of the set of additional keywords.
Malladi teaches wherein the semantically scoring comprises: assigning a weight to each of the set of additional keywords; and calculating a semantic score for each sentence based on the weight assigned to each of the set of additional keywords (see at least Col 29 lines 47 to Col 31 line 19: teaches scoring and assigning weights to keywords. Provides significance measure for candidate keywords used in generating summary for content. “keyword correlation graph data object for a multi-party communication transcript data object describes a graph having nodes/vertices that correspond to candidate keywords of the multi-party communication transcript data object and edges/links associated with edge/link weights that correspond to semantic correlations between candidate keyword pairs associated with node/vertex pairs of the graph”… “a keyword score is a computed/predicted semantic/summarization-related significance measure for a respective candidate keyword in a respective multi-party communication transcript data object that is generated based at least in part on the document-keyword similarity score for the candidate keyword and a local keyword correlation graph data object.”).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention these teachings of Malladi in combination with Kandekar in order to semantically score candidate keywords (associated keywords), use in the summary process, since it enhances the method for generating the summary by providing a more relevant, coherent, and information-rich summary.
Claim(s) 4 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kandekar et al. (US 20120210203), in view of Chhabra et al. (US 20200293618) and further in view of Saggi et al. (US 20200372066).
In regards to claim 4, Kandekar doesn’t specifically teach wherein: the contextual classification is one of a rule-based classification or a model- based classification, performing the contextual classification to select the set of sentences based on the rule-based classification comprises: for each unlabeled data, identifying the plurality of sentences having a rank greater than a predefined threshold; and selecting a set of sentences based on the identifying, wherein the set of sentences is selected with the rank greater than the predefined threshold; and performing the contextual classification to select the set of sentences based on the model-based classification comprises: for each labelled data, training a model-based classification model based on the one or more keywords and the additional set of keywords; and selecting a set of sentences using the model-based classification model.
Saggi teaches wherein: the contextual classification is one of a rule-based classification or a model- based classification (teaches model based contextual classification. See at least para 104:: “The context classification may include, for example, automatically assigning a category label to text, which the system may perform by using existing databases of text categories or by using learned and/or optimized text categories”), performing the contextual classification to select the set of sentences based on the rule-based classification comprises: for each unlabeled data, identifying the plurality of sentences having a rank greater than a predefined threshold; and selecting a set of sentences based on the identifying, wherein the set of sentences is selected with the rank greater than the predefined threshold; and performing the contextual classification to select the set of sentences based on the model-based classification comprises: for each labelled data, training a model-based classification model based on the one or more keywords and the additional set of keywords; and selecting a set of sentences using the model-based classification model (teaches model based contextual classification based on learn model for optimized categories. See at least para 104: “The context classification may include, for example, automatically assigning a category label to text, which the system may perform by using existing databases of text categories or by using learned and/or optimized text categories” On para 110: “generation of the importance sub-scores may include generating and training one or more machine learning models towards identifying the importance-determining features specific to each analytical modality.”).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention these teachings of Saggi in combination with Kandekar in order to incorporate model-based context classification, since it enhances the method for understanding and leveraging the semantic and structural context of the content to be summarized so as to provide a more relevant and coherent summary.
Claim(s) 5 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kandekar et al. (US 20120210203), in view of Chhabra et al. (US 20200293618) and further in view of Kunal Kandekar et al. (US 20120079372) (hereinafter Kunal).
In regards 5, Kandekar teaches clustering as taught above (See at least para 70, 76, 81), but doesn’t specially teach wherein clustering the set of sentences based on the semantic scoring comprises: identifying one or more sentences within the set of sentences having relevant semantic scores; and grouping each of the one or more sentences into a separate cluster based on the identifying, wherein each of the separate cluster comprises the one or more sentences with a corresponding relevant semantic scores.
Kunar teaches wherein clustering the set of sentences based on the semantic scoring comprises: identifying one or more sentences within the set of sentences having relevant semantic scores; and grouping each of the one or more sentences into a separate cluster based on the identifying, wherein each of the separate cluster comprises the one or more sentences with a corresponding relevant semantic scores (teaches clustering based on scoring of a semantic analysis. See at least para 33: “determine a similarity relationship (e.g., a semantic relationship) between the text seed and one or more adjacent text subgroups that do not include the identified text seed, to associate the text seed with the one or more adjacent text subgroups based on the similarity relationship to create a text cluster, and to link the text cluster to another one of the text subgroups within the electronic text based on a similarity relationship between the text cluster and the other one of the text subgroups.”…” The text seed may be selected based on a relevance or similarity score assigned to words of the electronic text. The similarity relationship between the text seed and adjacent text subgroups may be determined by, for example, but not limited to, performing semantic analysis”.)
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention these teachings of Kunal in combination with Kandekar in order to accomplished sentence clustering in a similar manner, since by applying these clustering steps it enhances the method by improving how the sentences are group and selected based on a determined relationship and thus providing a better summary to the user.
Claim(s) 6-7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kandekar et al. (US 20120210203), in view of Chhabra et al. (US 20200293618) and further in view of Elango et al. (US 20230090049).
In regards 6, Kandekar teaches clustering as taught above (See at least para 70, 76, 81), but doesn’t specially teach wherein clustering the set of sentences based on the embedding comprises: identifying one or more sentences within the set of sentences having relevant embeddings; and grouping each of the one or more sentences into a separate cluster based on the identifying, wherein each of the separate cluster comprises the one or more sentences with a corresponding relevant embeddings.
Elango teaches wherein clustering the set of sentences based on the embedding comprises: identifying one or more sentences within the set of sentences having relevant embeddings; and grouping each of the one or more sentences into a separate cluster based on the identifying, wherein each of the separate cluster comprises the one or more sentences with a corresponding relevant embeddings (teaches generating summaries of content using sentence embedding where the similar sentence which are relevant are clustered together. See at least para 103-107: “text summarization process, each cluster of sentence embeddings is interpreted as a set of semantically similar sentences having a meaning expressed by a single candidate sentence within the summary. The candidate sentence is chosen to be the sentence having a vector representation closest to the cluster center. Candidate sentences corresponding to each cluster are then ordered to form a summary”).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention these teachings of Elango in combination with Kandekar in order to accomplished sentence clustering in a similar manner using embedding, since by applying these clustering steps it enhances the method by improving how the sentences are grouped, then candidate sentence are selected thus providing a precise summary to the user.
In regards to claim 7, Kandekar teaches wherein generating the consolidated summarized text comprises: generating a summarized text for each of the separate cluster; and concatenating the summarized text of each of the separate cluster to obtain a consolidated summarized text (teaches generating summary based on summarized segments as the sentence have been clustered. See para 35-38, 53, 59: “summary generation for a given level of detail and a given user profile for a content segment may be performed by: traversing the different levels of the hierarchical summary segment for that content segment; matching the metadata of each level with the required detail and the user profile; identifying the level or levels having the best match; and selecting the summary text from the identified level. ”. Also see para 70, 76, 81: “cluster (step 606) and applying an automated summarization technique to the cluster (step 608). For example, the media summary generator 118 may access a first cluster for the e-book 116, which corresponds to chapters 1-5. A summarization technique may be applied to the content of chapters 1-5. For example, the media summary generator 118 may apply a semantic analysis technique, an NLP technique, a summary extraction technique, the like, or combinations thereof to the content of chapters 1-5. The resulting summarization of the content associated with the cluster may be referred to as a "summary segment”. On para 81-82: “Steps 606-618 may be performed iteratively until there are no more clusters. In response to determining that there are no more clusters at step 618, the method may proceed to step 620. At step 620 of FIG. 6, the method includes presenting the content summary).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Akolkar et al. (US 20190027188): teaches concise summarization of a media production is achieved using cognitive analysis which groups segments of the production into clusters based on extracted features, selects a representative segment for each cluster, and combines the representative segments to form a summary.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARIO M VELEZ-LOPEZ whose telephone number is (571)270-7971. The examiner can normally be reached on M-F 10:30am-5:30pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Scott Baderman, can be reached at telephone number 571-272-3644. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from Patent Center and the Private Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from Patent Center or Private PAIR. Status information for unpublished applications is available through Patent Center and Private PAIR for authorized users only. Should you have questions about access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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) Form at https://www.uspto.gov/patents/uspto-automated- interview-request-air-form.
/MARIO M VELEZ-LOPEZ/
Examiner, Art Unit 2118
/SCOTT T BADERMAN/Supervisory Patent Examiner, Art Unit 2118
