DETAILED ACTION
This communication is in response to the Amendments and Arguments filed on 02/24/2026.
Claim(s) 1-20 are pending and have been examined. Hence, this action has been made FINAL.
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 .
Response to Arguments and Amendments
Amendments to the claims by the Applicant have been considered and addressed below.
With respect to the Allowable Subject Matter, Drawing Objections, Claim Objections, 35 USC § 101, and 103 rejections, the Applicant provides several arguments in which the Examiner will respond accordingly, below.
Drawing Objection(s)
Arguments in page 8 of the Remarks filed on 02/24/2026
Examiner’s Response to Arguments:
Applicant’s arguments with respect to the drawing objections have been fully considered and are persuasive. The drawing objections of Figs. 2-6 have been withdrawn.
Claim Objection(s)
Arguments in page 9 of the Remarks filed on 02/24/2026
Examiner’s Response to Arguments:
Applicant’s arguments with respect to the claim objections have been fully considered and are persuasive. The claim objections of claims 6 and 17 have been withdrawn.
35 USC § 101 rejection(s)
Arguments in page 9-12 of the Remarks filed on 02/24/2026
Examiner’s Response to Arguments:
Arguments have been considered but these are not persuasive. The Examiner respectfully disagrees with the arguments of “the present claims are patent eligible at least at Step 2A Prong Two because the claims recite additional elements, and those elements integrate the abstract idea into a practical application as the claim improves the technical field of automated text processing, by providing a two-stage process for identifying key phrases in an input text: 1) extraction of candidate phrases; and 2) extraction of contextual features utilizing an enriched knowledge base.”, “Independent claims 1, 12, and 20 (as amended) recite additional elements that do not fall within any of the enumerated groupings of abstract ideas-mathematical concepts, certain methods of organizing human activity, or mental processes. ” and “Because the claims improve the technical field of automated text processing, the additional elements interact with and impact the remaining limitations so as to integrate the abstract idea into a practical application, pursuant to MPEP §§2106.04(d)(1) and 2106.05(a). Therefore, the claims are not directed to the judicial exception, and the claims are patent eligible at least at Step 2A Prong Two.”
The Examiner notes that under the broadest reasonable interpretation, the limitations are considered to be practically performed by a human (mentally and/or using pen and paper) and refers the Applicant to the following MPEP sections:
MPEP 2106.04(a)(2)(C)(III):
MENTAL PROCESSES
Nor do the courts distinguish between claims that recite mental processes performed by humans and claims that recite mental processes performed on a computer. As the Federal Circuit has explained, "[c]ourts have examined claims that required the use of a computer and still found that the underlying, patent-ineligible invention could be performed via pen and paper or in a person’s mind." Versata Dev. Group v. SAP Am., Inc., 793 F.3d 1306, 1335, 115 USPQ2d 1681, 1702 (Fed. Cir. 2015). See also Intellectual Ventures I LLC v. Symantec Corp., 838 F.3d 1307, 1318, 120 USPQ2d 1353, 1360 (Fed. Cir. 2016) (‘‘[W]ith the exception of generic computer-implemented steps, there is nothing in the claims themselves that foreclose them from being performed by a human, mentally or with pen and paper.’’); Mortgage Grader, Inc. v. First Choice Loan Servs. Inc., 811 F.3d 1314, 1324, 117 USPQ2d 1693, 1699 (Fed. Cir. 2016) (holding that computer-implemented method for "anonymous loan shopping" was an abstract idea because it could be "performed by humans without a computer"). Mental processes recited in claims that require computers are explained further below with respect to point C.
MPEP 2106.04(a)(2)(C)(III)(C):
A Claim That Requires a Computer May Still Recite a Mental Process
Claims can recite a mental process even if they are claimed as being performed on a computer. The Supreme Court recognized this in Benson, determining that a mathematical algorithm for converting binary coded decimal to pure binary within a computer’s shift register was an abstract idea. The Court concluded that the algorithm could be performed purely mentally even though the claimed procedures "can be carried out in existing computers long in use, no new machinery being necessary." 409 U.S at 67, 175 USPQ at 675. See also Mortgage Grader, 811 F.3d at 1324, 117 USPQ2d at 1699 (concluding that concept of "anonymous loan shopping" recited in a computer system claim is an abstract idea because it could be "performed by humans without a computer").
In evaluating whether a claim that requires a computer recites a mental process, examiners should carefully consider the broadest reasonable interpretation of the claim in light of the specification. For instance, examiners should review the specification to determine if the claimed invention is described as a concept that is performed in the human mind and applicant is merely claiming that concept performed 1) on a generic computer, or 2) in a computer environment, or 3) is merely using a computer as a tool to perform the concept. In these situations, the claim is considered to recite a mental process. )
Please see detailed analysis below (Prong Two) for more details on how the Examiner understands the independent claims do not recite additional elements that integrate the judicial exception into a practical application. Hence, not qualifying as patent eligible subject matter under 35 U.S.C. § 101.
Please refer to MPEP 2106.04(II): Eligibility Step 2A: Whether a Claim is Directed to a Judicial Exception: (A) Step 2A is a Two-Prong Inquiry:
(1) Prong One:
Prong One asks does the claim recite an abstract idea, law of nature, or natural phenomenon? In Prong One examiners evaluate whether the claim recites a judicial exception, i.e. whether a law of nature, natural phenomenon, or abstract idea is set forth or described in the claim. While the terms "set forth" and "described" are thus both equated with "recite", their different language is intended to indicate that there are two ways in which an exception can be recited in a claim. For instance, the claims in Diehr, 450 U.S. at 178 n. 2, 179 n.5, 191-92, 209 USPQ at 4-5 (1981), clearly stated a mathematical equation in the repetitively calculating step, and the claims in Mayo, 566 U.S. 66, 75-77, 101 USPQ2d 1961, 1967-68 (2012), clearly stated laws of nature in the wherein clause, such that the claims "set forth" an identifiable judicial exception. Alternatively, the claims in Alice Corp., 573 U.S. at 218, 110 USPQ2d at 1982, described the concept of intermediated settlement without ever explicitly using the words "intermediated" or "settlement." […]
An example of a claim that recites a judicial exception is "A machine comprising elements that operate in accordance with F=ma." This claim sets forth the principle that force equals mass times acceleration (F=ma) and therefore recites a law of nature exception. Because F=ma represents a mathematical formula, the claim could alternatively be considered as reciting an abstract idea. Because this claim recites a judicial exception, it requires further analysis in Prong Two in order to answer the Step 2A inquiry. An example of a claim that merely involves, or is based on, an exception is a claim to "A teeter-totter comprising an elongated member pivotably attached to a base member, having seats and handles attached at opposing sides of the elongated member." This claim is based on the concept of a lever pivoting on a fulcrum, which involves the natural principles of mechanical advantage and the law of the lever. However, this claim does not recite these natural principles and therefore is not directed to a judicial exception (Step 2A: NO). Thus, the claim is eligible at Pathway B without further analysis.
From this analysis, in Step 2A, Prong One, the Examiner has evaluated the independent claims accordingly and determined that the amended independent claims as drafted indeed describe a judicial exception (i.e., an abstract idea), which represent a mental process (which can be performed by a human with pen and paper).
More specifically, similar to what was discussed in the Non-Final Rejection mailed on 12/03/2025:
The independent claim(s) recite(s):
1. A system comprising:
at least one processor; and
at least one memory storing instructions that, when executed by the at least one processor, cause operations comprising:
extracting a plurality of candidate phrases from a first field of a textual input;
extracting first context data from a second field of the textual input;
querying an external database to retrieve second context data related to the textual input, the external database comprising a knowledge base with potential solutions to user problems;
combining the first context data and the second context data to form combined context data;
executing a sentence transformer model having been trained to generate vectorized versions of the plurality of candidate phrases and the combined context data;
storing the vectorized versions of the plurality of candidate phrases and the combined context data;
determining, for each candidate phrase of the plurality of candidate phrases, a similarity score between a vectorized version of the candidate phrase and a vectorized version of the combined context data;
selecting a subset of candidate phrases from the plurality of candidate phrases, wherein the subset of candidate phrases have highest determined similarity scores of a plurality of determined similarity scores corresponding to the plurality of candidate phrases; and
providing the subset of candidate phrases to one or more applications which are generating responses to the textual input.
12. A method comprising:
[the limitations as in claim 1, above.]
20. A non-transitory computer readable medium storing instructions, which when executed by at least one data processor, result in operations comprising:
[the limitations as in claim 1, above.]
This reads on a human (e.g., mentally and/or using pen and paper):
Reading text and identifying phrases in a field of said text (e.g., from a form or document);
identifying further text (e.g., contextual information) in a second field of said text (e.g., from a form or document);
reading a second textual source (e.g., dictionary, newspaper, etc.) to get further contextual information;
consider or combine both contextual information (i.e., from the form or document and the second textual source);
using predetermined set of rules to re-write the phrases previously identified as well as the contextual information following a predefined set of steps (e.g., assigning numerical values);
writing down said phrases / values;
determining or calculating (e.g., mathematical concept) a similarity between the re-written phrases and the re-written contextual information;
selecting the phrase(s) with the highest similarity; and
writing down the selected phrase.
Please also refer to MPEP 2106.05(f)(2): Whether the claim invokes computers or other machinery merely as a tool to perform an existing process, and MPEP 2106.06(b): Clear Improvement to a Technology or to Computer Functionality.
Please refer to MPEP 2106.04(II): Eligibility Step 2A: Whether a Claim is Directed to a Judicial Exception: (A) Step 2A is a Two-Prong Inquiry:
(2) Prong Two:
Prong Two asks does the claim recite additional elements that integrate the judicial exception into a practical application? In Prong Two, examiners evaluate whether the claim as a whole integrates the exception into a practical application of that exception. If the additional elements in the claim integrate the recited exception into a practical application of the exception, then the claim is not directed to the judicial exception (Step 2A: NO) and thus is eligible at Pathway B. This concludes the eligibility analysis. If, however, the additional elements do not integrate the exception into a practical application, then the claim is directed to the recited judicial exception (Step 2A: YES), and requires further analysis under Step 2B (where it may still be eligible if it amounts to an ‘‘inventive concept’’). For more information on how to evaluate whether a judicial exception is integrated into a practical application, see MPEP § 2106.04(d)(2).
From this analysis, in Step 2A, Prong Two, the Examiner has evaluated the independent claims accordingly and determined that the amended independent claims as drafted that the claims as a whole do not include additional elements that integrate the exception into a practical application of that exception. (i.e., an abstract idea). As discussed in the Non-Final Rejection mailed on 12/03/2025:
This judicial exception is not integrated into a practical application because for example: claims 1 and 20 recite: system, processor, memory, non-transitory computer readable medium, external database, knowledge base, sentence transformer model and/or one or more applications. As an example, in ¶ [0068] of the as filed specification, it is disclosed that “The systems and methods disclosed herein can be embodied in various forms including, for example, a data processor, such as a computer that also includes a database, digital electronic circuitry, firmware, software, or in combinations of them. Moreover, the above-noted features and other aspects and principles of the present disclosed implementations can be implemented in various environments. Such environments and related applications can be specially constructed for performing the various processes and operations according to the disclosed implementations or they can include a general-purpose computer or computing platform selectively activated or reconfigured by code to provide the necessary functionality. The processes disclosed herein are not inherently related to any particular computer, network, architecture, environment, or other apparatus, and can be implemented by a suitable combination of hardware, software, and/or firmware. For example, various general-purpose machines can be used with programs written in accordance with teachings of the disclosed implementations, or it can be more convenient to construct a specialized apparatus or system to perform the required methods and techniques.” Therefore, a general-purpose computer or computing device is described and mainly used as an application thereof. Accordingly, these additional elements do not integrate the abstract idea into a practical idea because it does not impose any meaningful limits on practicing the abstract idea.
Please also refer to MPEP 2106.05(f)(2): Whether the claim invokes computers or other machinery merely as a tool to perform an existing process.
Finally, please refer to MPEP 2106.05(A): Relevant Considerations For Evaluating Whether Additional Elements Amount To An Inventive Concept
Limitations that the courts have found not to be enough to qualify as "significantly more" when recited in a claim with a judicial exception include:
i. Adding the words "apply it" (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, e.g., a limitation indicating that a particular function such as creating and maintaining electronic records is performed by a computer, as discussed in Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984 (see MPEP § 2106.05(f));
ii. Simply appending well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception, e.g., a claim to an abstract idea requiring no more than a generic computer to perform generic computer functions that are well-understood, routine and conventional activities previously known to the industry, as discussed in Alice Corp., 573 U.S. at 225, 110 USPQ2d at 1984 (see MPEP § 2106.05(d));
From this analysis, in Step 2B, the Examiner has evaluated the independent claims accordingly and determined that the independent claims as drafted have limitations that the courts have found not to be enough to qualify as "significantly more" when recited in a claim with a judicial exception. Similar to what was discussed in the Non-Final Rejection mailed on 12/03/2025:
The claim(s) does/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 elements of using a computer is listed as a general computing device as noted. The claim is not patent eligible.
For more details, please refer to the 35 U.S.C. § 101 rejections for claims 1-20, below.
35 USC § 103 rejection(s)
Arguments in page 12-15 of the Remarks filed on 02/24/2026
Examiner’s Response to Arguments:
Applicant’s arguments with respect to claim(s) 1, 12, and 20 under 35 U.S.C. § 103 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.
Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Cai et al. (US 20190347282 A1) and further in view of Yang et al. (US 20230091076 A1) and Duboue et al. (US 20120078902 A1).
For more details, please refer to updated 35 U.S.C. § 103 rejections for claims 1, 12, and 20, below.
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.
Claim(s) 1-20 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. More specifically directed to the abstract idea grouping of: mental process and/or mathematical concept.
The independent claim(s) recite(s):
1. A system comprising:
at least one processor; and
at least one memory storing instructions that, when executed by the at least one processor, cause operations comprising:
extracting a plurality of candidate phrases from a first field of a textual input;
extracting first context data from a second field of the textual input;
querying an external database to retrieve second context data related to the textual input, the external database comprising a knowledge base with potential solutions to user problems;
combining the first context data and the second context data to form combined context data;
executing a sentence transformer model having been trained to generate vectorized versions of the plurality of candidate phrases and the combined context data;
storing the vectorized versions of the plurality of candidate phrases and the combined context data;
determining, for each candidate phrase of the plurality of candidate phrases, a similarity score between a vectorized version of the candidate phrase and a vectorized version of the combined context data;
selecting a subset of candidate phrases from the plurality of candidate phrases, wherein the subset of candidate phrases have highest determined similarity scores of a plurality of determined similarity scores corresponding to the plurality of candidate phrases; and
providing the subset of candidate phrases to one or more applications which are generating responses to the textual input.
12. A method comprising:
[the limitations as in claim 1, above.]
20. A non-transitory computer readable medium storing instructions, which when executed by at least one data processor, result in operations comprising:
[the limitations as in claim 1, above.]
This reads on a human (e.g., mentally and/or using pen and paper):
Reading text and identifying phrases in a field of said text (e.g., from a form or document);
identifying further text (e.g., contextual information) in a second field of said text (e.g., from a form or document);
reading a second textual source (e.g., dictionary, newspaper, etc.) to get further contextual information;
consider or combine both contextual information (i.e., from the form or document and the second textual source);
using predetermined set of rules to re-write the phrases previously identified as well as the contextual information following a predefined set of steps (e.g., assigning numerical values);
writing down said phrases / values;
determining or calculating (e.g., mathematical concept) a similarity between the re-written phrases and the re-written contextual information;
selecting the phrase(s) with the highest similarity; and
writing down the selected phrase.
This judicial exception is not integrated into a practical application because for example: claims 1 and 20 recite: system, processor, memory, non-transitory computer readable medium, external database, knowledge base, sentence transformer model and/or one or more applications. As an example, in ¶ [0068] of the as filed specification, it is disclosed that “The systems and methods disclosed herein can be embodied in various forms including, for example, a data processor, such as a computer that also includes a database, digital electronic circuitry, firmware, software, or in combinations of them. Moreover, the above-noted features and other aspects and principles of the present disclosed implementations can be implemented in various environments. Such environments and related applications can be specially constructed for performing the various processes and operations according to the disclosed implementations or they can include a general-purpose computer or computing platform selectively activated or reconfigured by code to provide the necessary functionality. The processes disclosed herein are not inherently related to any particular computer, network, architecture, environment, or other apparatus, and can be implemented by a suitable combination of hardware, software, and/or firmware. For example, various general-purpose machines can be used with programs written in accordance with teachings of the disclosed implementations, or it can be more convenient to construct a specialized apparatus or system to perform the required methods and techniques.” Therefore, a general-purpose computer or computing device is described and mainly used as an application thereof. Accordingly, these additional elements do not integrate the abstract idea into a practical idea because it does not impose any meaningful limits on practicing the abstract idea.
The claim(s) does/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 elements of using a computer is listed as a general computing device as noted. The claim is not patent eligible.
With respect to claims 2 and 13, the claim(s) recite:
2 and 13. The system/method of claims 1 and 12, wherein the similarity score is calculated as a cosine similarity between the vectorized version of the candidate phrase and the vectorized version of the combined context data.
This reads on a human (e.g., mentally and/or using pen and paper):
determining or calculating (e.g., mathematical concept) a similarity (e.g., cosine similarity) between the re-written phrases and the re-written contextual information
No additional limitations are present.
With respect to claims 3 and 14, the claim(s) recite:
3 and 14. The system/method of claims 1 and 12, wherein the plurality of candidate phrases are extracted from the first field of the textual input using a graph ranking method.
This reads on a human (e.g., mentally and/or using pen and paper):
extracting the phrases using a predetermined set of rules (e.g., graph-based ranking)
No additional limitations are present.
With respect to claims 4 and 15, the claim(s) recite:
4 and 15. The system/method of claims 1 and 12, wherein the textual input is a customer ticket.
This reads on a human (e.g., mentally and/or using pen and paper):
analyzing a customer ticket (e.g., received text)
No additional limitations are present.
With respect to claims 5 and 16, the claim(s) recite:
5 and 16. The system/method of claims 4 and 15, wherein the first field is a title of the customer ticket.
This reads on a human (e.g., mentally and/or using pen and paper):
wherein the customer ticket (e.g., received text) includes a title
No additional limitations are present.
With respect to claims 6 and 17, the claim(s) recite:
6 and 17. The system/method of claims 5 and 16, wherein the second field is a problem description field of the customer ticket.
This reads on a human (e.g., mentally and/or using pen and paper):
wherein the customer ticket (e.g., received text) includes a problem description
No additional limitations are present.
With respect to claims 7 and 18, the claim(s) recite:
7 and 18. The system/method of claims 1 and 12, wherein extracting the plurality of candidate phrases from the first field of the textual input comprises:
extracting a plurality of potential candidate phrases from the first field of the textual input;
scoring each potential candidate phrase of the plurality of potential candidate phrases based at least on: a presence of technical terms in the potential candidate phrase, if the potential candidate phrase is capitalized, and a frequency of words of the potential candidate phrase in the first field of the textual input, the second field of the textual input, and one or more secondary data sources; and
selecting a subset of the plurality of potential candidate phrases having highest scores out of the plurality of potential candidate phrases.
This reads on a human (e.g., mentally and/or using pen and paper):
Reading text and identifying phrases in a field of said text (e.g., from a form or document);
Assigning a score to the phrases based on a predefined set of rules associated with: terminology used (e.g., technical), capitalization of words, frequency of words in the text and textual sources; and
Selecting the phrase with the highest score.
No additional limitations are present.
With respect to claims 8 and 19, the claim(s) recite:
8 and 19. The system/method of claims 1 and 12, wherein the operations further comprise:
generating a first score for each potential candidate phrase based on the presence of technical terms in the potential candidate phrase;
generating a second score for each potential candidate phrase based on if the potential candidate phrase is capitalized; and
generating a third score for each potential candidate phrase based on how frequently words of the potential candidate phrase appear in the textual input and in the one or more data sources related to the textual input.
This reads on a human (e.g., mentally and/or using pen and paper):
Assigning a score to the phrases based on a predefined set of rules associated with: terminology used (e.g., technical);
Assigning a score to the phrases based on a predefined set of rules associated with: capitalization of words; and
Assigning a score to the phrases based on a predefined set of rules associated with: frequency of words in the text and textual sources.
No additional limitations are present.
With respect to claim 9, the claim(s) recite:
9. The system of claim 8, wherein the operations further comprise combining the first score, the second score, and the third score to generate a combined score for each potential candidate phrase.
This reads on a human (e.g., mentally and/or using pen and paper):
Combining the scores defined above using a predetermined set of rules.
No additional limitations are present.
With respect to claim 10, the claim(s) recite:
10. The system of claim 9, wherein the operations further comprise sorting the plurality of potential candidate phrases based on corresponding combined scores.
This reads on a human (e.g., mentally and/or using pen and paper):
Ordering the phrases based on the combination of scores.
No additional limitations are present.
With respect to claim 11, the claim(s) recite:
11. The system of claim 10, wherein the operations further comprise selecting one or more potential candidate phrases having highest combined scores as key phrases of the textual input.
This reads on a human (e.g., mentally and/or using pen and paper):
Selecting the phrase with the highest score.
No additional limitations are present.
Claim Rejections - 35 USC § 103
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-6, 12-17 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Cai et al. (US 20190347282 A1) and further in view of Yang et al. (US 20230091076 A1) and Duboue et al. (US 20120078902 A1).
As to independent claim 1, Cai et al. teaches:
1. A system (see ¶ [0009]: “Systems, methods, and computer readable media storing machine interpretable instructions are described in various embodiments for receiving string inputs representative of a new incident ticket and generating an output data structure representative of a subset of potential solutions selected from a plurality of potential solutions.”) comprising:
at least one processor (see ¶ [0031]: “In accordance with an aspect, there is provided an incident management platform for incident ticket volume prediction having a processor and a memory storing machine executable instructions to configure the processor to:…”) ; and
at least one memory storing instructions (see ¶ [0031] citation as in limitation above.) that, when executed by the at least one processor, cause operations (see ¶ [0031] citation as in limitation above.) comprising:
extracting a plurality of candidate phrases from a first field of a textual input (see ¶ [0009] citation as in limitation above and further ¶ [0011-0012]: “[0011] The data receiver is configured for receiving the string inputs representative of the new incident ticket, the new incident ticket including at least a title field and a description field. [0012] A pre-processing engine is configured to concatenate the title field and the description field of the new incident ticket, and to remove low value words from the concatenated title field and description field. Low value words can be identified, for example, as stop words based on a natural language processing library.”
and ¶ [0077]: “The communication interface 106, in some embodiments, is a data receiver configured for receiving the string inputs representative of the new incident ticket, the new incident ticket including at least a title field (e.g., database error) and a description field (e.g., “How do I fix inc0304036”).”);
extracting first context data from a second field of the textual input (see ¶ [0009] citation as in limitation above and further ¶ [0011-0012 and 0077]: “[0011] The data receiver is configured for receiving the string inputs representative of the new incident ticket, the new incident ticket including at least a title field and a description field. [0012] A pre-processing engine is configured to concatenate the title field and the description field of the new incident ticket, and to remove low value words from the concatenated title field and description field. Low value words can be identified, for example, as stop words based on a natural language processing library.”
and ¶ [0077]: “The communication interface 106, in some embodiments, is a data receiver configured for receiving the string inputs representative of the new incident ticket, the new incident ticket including at least a title field (e.g., database error) and a description field (e.g., “How do I fix inc0304036”).”);
selecting a subset of candidate phrases from the plurality of candidate phrases (see ¶ [0009 and 0011-0012] citations as in limitations above and further ¶ [0079]: “The processor 104 is configured to provide a pre-processing engine configured to concatenate the title field and the description field of the new incident ticket, and to remove low value words from the concatenated title field and description field. The pre-processing is adapted to aid in ensuring that comparable vectors can be formed from the incident ticket string, especially after resolutions are identified and later incorporated into the incident-solution pair repository. In some embodiments, the vectors are extended with additional fields for derivative or metadata-based data elements.”
and ¶ [0164]: “The natural language processor 120 can include instructions or scripts which can include a text similarity process 602 (e.g., Term Frequency and Cosine Similarity) that can involve the following operations:
Concatenate Title and Description of new incident ticket;
Remove stop words from the title/description;
Performs Term Frequency Inverse Document Frequency on Knowledge base 608 (of data storage 110) new Incident with ngram range (1,3);
Execute cosine similarity on the new incident title/description against all historical incidents;
Sort the output of cosine similarity based on closet matching incident tickets; Determine top 3 (or more) indexes and map them back to their respective incident numbers;
Output the incident numbers and confidence scores.”),
wherein the subset of candidate phrases have highest determined similarity scores of a plurality of determined similarity scores corresponding to the plurality of candidate phrases (see ¶ [0009 and 0011-0012] citations as in limitations above as well as ¶ [0079 and 0164] citations as in limitation above (i.e., removal of low value words (e.g., stop words from the title – candidate phrases))
and further ¶ [0085-0086, 0090-0091, and 0093]: “[0085] As the string corresponding to the new incident is likely not an exact match to any historical incident, both the new incident string and historical strings can be converted into vector representations. Term frequency inverse document frequency analysis is performed across against a set of historical incident-solution pairs stored on a data repository and the pre-processed concatenated title field and description field to establish an initial set of vectors for analysis. [0086] Term frequency inverse document frequency analysis is a weighted word process which helps in identifying word patterns (bigrams and trigrams) from a document/text. N-gram is a hyper parameter which is used in the term frequency inverse document frequency analysis approach to specify the word count range. N-gram range(1,3) specifies up to 3 word combination for the TFIDF. [0090] The natural language processor 120 is configured to determine a plurality of cosine similarity scores, each corresponding to the pre-processed concatenated title field and description field mapped against a historical incident-solution pair stored on the data repository. The cosine similarity scores are established as a measure of similarity between two non-zero vectors of an inner product space. [0091] Cosine similarity is a distance approach which determines which strings are most closely related. The above results are plotted in a multi-dimensional space graph based on the n-grams, establishing a constellation of points. In some embodiments, a number of incidents which are the closest are selected as potentially relevant historical incident-solution pairs selected as a subset from all of the historical incident-solution pairs. [0093] Accordingly, a data structure (e.g., a database table) can be generated where, for a specific new incident, a plurality of historical incident tickets (and their corresponding solutions as incident-solution pairs) can each be assigned a similarity score and in some embodiments, a confidence score for a potential match. The data structure elements may be sorted by the similarity score and used to generate an ordered list of closest matching historical incidents of the set of historical incident-solution pairs stored on the data repository, ordered by the determined cosine similarity scores.”); and
providing the subset of candidate phrases to one or more applications which are generating responses to the textual input (see ¶ [0009, 0011-0012, 0079, 0085-0086, 0090-0091, 0093, and 0164] citations as in limitations above
and further ¶ [0094-0096]: “[0094] The natural language processor 120 determines the subset of potential solutions selected from the plurality of potential solutions based on a pre-defined number of closest matching historical incidents from the ordered list of the closest matching historical incidents, which are then utilized to generate the output data structure representative of the subset of potential solutions selected from the plurality of potential solutions. [0095] The prediction/prescriptive models 126, in some embodiments, can then utilize the output data structure to identify one or more prescriptive actions. The one or more prescriptive actions can include solution strings which can be interpreted by an agent who then takes corrective action (e.g., by selecting to initiate a corresponding data process). [0096] In some embodiments, the processor 104 is configured to output the prescriptive solution to a virtual agent 180 (FIG. 2). In this example, the virtual agent 180 may then provide a graphical user interface or other output to indicate to the user what a potential solution is, and the data structure of potential solutions can be traversed to provide potential alternate solutions, provided such solutions have confidence scores that are above a pre-defined threshold. For example, the virtual agent 180 may receive an incident string “What to do for error code 0x22FFFE?”, and upon traversing historical incident tickets, respond with “There have been three successful resolutions of similar problems. The most likely solution is to allocate additional memory to the page file”.”) .
However, Cai et al. does not explicitly teach, but Yang et al. does teach:
combining the first context data and the second context data to form combined context data (see ¶ [0007, 0011, and 0046-0047]: ¶ [0007 and 0011]: “[0007] Example 1 is a computer-implemented method for extracting a set of keyphrases from an input text, the computer-implemented method comprising: receiving the input text having a plurality of words; identifying, using at least one unsupervised machine-learning model, a set of candidate phrases from the plurality of words, wherein each of the set of candidate phrases includes one or more words from the plurality of words; and selecting, using at least one supervised machine-learning model, the one or more keyphrases from the set of candidate phrases, the at least one supervised machine-learning model having been previously trained using a set of training examples. [0011] Example 5 is the computer-implemented method of example(s) 1-4, wherein the input text includes at least one of a title, a facet, a short description, or a long description.”
and ¶ [0046-0047]: “[0046] Some embodiments of the present disclosure include a data preprocessing step, in which an input text can filtered, denoised, refined, or otherwise modified to prepare for subsequent steps. In some instances, the input text may be a collection having a long description that includes multiple paragraphs. The long description might contain relevant and/or irrelevant information. In order to decrease noise for the task of candidate phrase identification as well as other tasks, it may be desirable to keep only the relevant information. In some instances, a collection input text might also have a short description that describes the most important information in the collection. [0047] In some instances, the short description may be expanded in the long description. To keep only the relevant information, in some embodiments, a Universal Sentence Encoder may be used to calculate the semantic similarity of the short description with each paragraph in the long description. The Universal Sentence Encoder may return an embedding vector for the short description and for each paragraph of the long description and/or other component of a collection or text, and the similarity between the short description and each paragraph may be calculated using an inner product of the embedding vectors. The paragraphs with the highest similarity (or greater than a predetermined or other threshold) are considered to be the most relevant and may be retrieved and retained as the main collection description.”
Here, the short description and long description are interpreted to be associated with the first context data and the second context data disclosed in the instant application.”);
executing a sentence transformer model having been trained to generate vectorized versions of the plurality of candidate phrases and the combined context data (see ¶ [0007, 0011, and 0046-0047] citations as in limitation above. More specifically: ¶ [0007]: “…identifying, using at least one unsupervised machine-learning model, a set of candidate phrases from the plurality of words, wherein each of the set of candidate phrases includes one or more words from the plurality of words; and selecting, using at least one supervised machine-learning model, the one or more keyphrases from the set of candidate phrases, the at least one supervised machine-learning model having been previously trained using a set of training examples.” and ¶ [0047]: “…To keep only the relevant information, in some embodiments, a Universal Sentence Encoder may be used to calculate the semantic similarity of the short description with each paragraph in the long description. The Universal Sentence Encoder may return an embedding vector for the short description and for each paragraph of the long description and/or other component of a collection or text, and the similarity between the short description and each paragraph may be calculated using an inner product of the embedding vectors...”
and further Fig. 6A (inserted below for reference with some highlighted areas/regions added. For Example: region #1 (red): title 630, region #2 (blue): short description 634 (includes the title), and region #3 (green): long description 636.
Here, the Examiner notes that the vectorization of the plurality of candidate phrases and the combined context data is read by Yang’s teachings of vectorizing or embedding the short description (including the title), the long description, and other text in the input text. Also, the Examiner notes that the sentence encoder returning an embedding vector for the short description and for each paragraph of the long description read on the sentence transformer model generating the vectorized vectors.)
storing the vectorized versions of the plurality of candidate phrases and the combined context data (see ¶ [0007, 0011, and 0046-0047] citations as in limitation above. More specifically: ¶ [0047]: “…The Universal Sentence Encoder may return an embedding vector for the short description and for each paragraph of the long description and/or other component of a collection or text, and the similarity between the short description and each paragraph may be calculated using an inner product of the embedding vectors…”
Here, the Examiner notes that the vectorization of the plurality of candidate phrases and the combined context data is read by Yang’s teachings of vectorizing or embedding the short description (including the title), the long description, and other text in the input text.);
determining, for each candidate phrase of the plurality of candidate phrases, a similarity score between a vectorized version of the candidate phrase and a vectorized version of the combined context data (see Fig. 6A and ¶ [0007, 0011, and 0046-0047] citations as in limitation above.
and further ¶ [0065-0066 and 0076-0077]: “[0065] FIGS. 6A-6I illustrate an example of an extraction of a set of keyphrases 620 from an input text 610 using a keyphrase extraction system, in accordance with some embodiments of the present disclosure. FIG. 6A shows an input text 610 having a plurality of words 612. In the illustrated example, words 612 are divided between a title 630, a set of facets 632, a short description 634, and a long description 636. The example illustrated in FIGS. 6A-6I corresponds to a genealogical description found in a genealogy database. [0066] FIG. 6B shows a data preprocessing step, performed by a data preprocessor, in which certain paragraphs and words 612 of input text 610 are removed to produce a processed input text 610A. In the illustrated example, each paragraph of long description 636 is compared to short description 634 to determine a similarity score, with higher similarity scores corresponding to higher levels of similarity. Based on the similarity score, it is determined whether each paragraph is selected or removed from input text 610. In the illustrated example, paragraphs with similarity scores less than 0.5 are removed while the only paragraph having a similarity score greater than 0.5 is selected and included in processed input text 610A. In some embodiments, paragraph similarity may be determined using a sentence encoder model. In addition, or alternatively, short text such as the title and/or facets may be pruned of irrelevant text. For example, in some embodiments TF-IDF can be used to prune irrelevant words from short text components of an input text such as “Golden Records,” “Free Access,” and other words in facets or otherwise that are not pertinent or informative of the topics of the collection. [0076] At step 702, an input text (e.g., input texts 110, 210, 610) having a plurality of words (e.g., words 112, 612) is received. The input text may include different sections, and the plurality of words may be divided between the sections. The sections may include a title (e.g., titles 230, 630), a set of facets (e.g., facets 232, 632), a short description (e.g., short descriptions 234, 634), and/or a long description (e.g., long descriptions 236, 636). The input text may be received by a keyphrase extraction system (e.g., keyphrase extraction systems 100, 200). [0077] At step 704, the input text is preprocessed by removing or modifying at least one of the plurality of words. In some embodiments, one or more sections of the input text may be removed. In some embodiments, each of the paragraphs of the long description may be compared to the short description to determine a similarity score for each paragraph, and paragraphs having similarity scores below a threshold may be removed. The input text may be preprocessed by a data preprocessor (e.g., data preprocessor 202) of the keyphrase extraction system.”);
PNG
media_image1.png
527
731
media_image1.png
Greyscale
Cai et al. and Yang et al. are considered to be analogous to the claimed invention because they are in the same field of endeavor in natural language data processing (e.g., entity/topic identification). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cai et al. to incorporate the teachings of Yang et al. of retrieving second context data from one or more data sources related to the textual input; combining the first context data and the second context data to form combined context data; vectorizing the plurality of candidate phrases and the combined context data; determining, for each candidate phrase of the plurality of candidate phrases, a similarity score between a vectorized version of the candidate phrase and a vectorized version of the combined context data which provides the benefit of having superior performance for the task of keyphrase extraction ([0045] of Yang et al.).
However, Cai et al. in combination with Yang et al. do not explicitly teach, but Duboue et al. does teach:
querying an external database to retrieve second context data related to the textual input, the external database comprising a knowledge base with potential solutions to user problems (see ¶ [0035-0036]: “[0035] As described below with respect to FIG. 3, after processing an input query to determine a LAT and searching and obtaining one or more candidate answers, there is performed for each candidate answer received the steps of matching the candidate against instances in a database which results in generating an output data structure, including the matched instances 132a; looking (searching) for evidence that the candidate answer has the required LAT and retrieving LT(s) associated with those instances in the knowledge base (KB) 132b; and attempting to match LAT(s) with LT(s) (e.g., as determined by a matching function that using a parser, a semantic interpreter and/or a simple pattern matcher) and producing a score representing the degree of match 132c. More particularly, the candidate answer LT and query LAT(s) are represented as lexical strings. Production of the score, referred to as a “TyCor” (Type Coercion) score, is comprised of the three steps: (1) candidate answer to instance matching, (2) instance to type association extraction, and (3) LAT to type matching. The score reflects the degree to which the candidate may be “coerced” to the LAT, where higher scores indicate a better coercion.
[0036] In one embodiment, as will be described herein with respect to FIG. 4, the present disclosure extends and complements the effectiveness of the system and method described in co-pending U.S. patent application Ser. No. 12/126,642 by automatically providing a source of information that associates “entities”, e.g., candidate answers for questions, to lexical types. Programmed components build or populate a repository of information, e.g., a database or knowledge base (KB) that can be used to accomplish the task of computing one or more lexical types (LT) for each candidate answer by facilitating automatic retrieval of “types” associated with answer instances (answer-typing) in a KB as described with respect to step 132b in FIG. 3. That is, given an instance (e.g., a word such as a noun) the method automatically evaluates the LT specified where the answer-typing data exists in a form that has a limited amount of explicit structure, i.e. semi-structured. In one aspect, the system and method produces a knowledge base of instances and types used in matching.”);
Cai et al., Yang et al., and Duboue et al. are considered to be analogous to the claimed invention because they are in the same field of endeavor in natural language data processing (e.g., entity/topic identification). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cai et al. in combination with Yang et al. to incorporate the teachings of Duboue et al. of querying an external database to retrieve second context data related to the textual input, the external database comprising a knowledge base with potential solutions to user problems which provides the benefit of providing a dynamic infrastructure and methodology for conducting question answering with deferred type evaluation using text with limited structure ([0012] of Duboue et al.).
As to independent claim 12, Cai et al. in combination with Yang et al. and Duboue et al. teaches the limitations as in claim 1, above.
Cai et al. further teaches:
12. A method (see ¶ [0009]: “Systems, methods, and computer readable media storing machine interpretable instructions are described in various embodiments for receiving string inputs representative of a new incident ticket and generating an output data structure representative of a subset of potential solutions selected from a plurality of potential solutions.”) comprising:
[the limitations as in claim 1, above.]
As to independent claim 20, Cai et al. in combination with Yang et al. and Duboue et al. teaches the limitations as in claim 1, above.
Cai et al. further teaches:
20. A non-transitory computer readable medium storing instructions (see ¶ [0009]: “Systems, methods, and computer readable media storing machine interpretable instructions are described in various embodiments for receiving string inputs representative of a new incident ticket and generating an output data structure representative of a subset of potential solutions selected from a plurality of potential solutions.”
and further ¶ [0249]: “The technical solution of embodiments may be in the form of a software product. The software product may be stored in a non-volatile or non-transitory storage medium, which can be a compact disk read-only memory (CD-ROM), a USB flash disk, or a removable hard disk…”), which when executed by at least one data processor, result in operations comprising:
[the limitations as in claim 1, above.]
Regarding claims 2 and 13, Cai et al. in combination with Yang et al.and Duboue et al. teaches the limitations as in claim 1 and 12, above.
Cai et al. further teaches:
2 and 13. The system/method of claims 1 and 12,
wherein the similarity score is calculated as a cosine similarity between the (see ¶ [0007]: “The natural language processing techniques include term frequency inverse document frequency analysis and the determination of cosine similarity scores, which are generated against each of a plurality of historical incident-solution pairs stored on the data repository. The cosine similarity scores are used for establishing indices for ordering/ranking the historical incident-solution pairs (e.g., based on a level of similarity to the new incident ticket), and in some embodiments, to establish confidence scores for each of the historical incident-solution pairs.”).
Yang et al. further teaches:
wherein the similarity score is calculated (see Fig. 6A and ¶ [0007, 0011, 0046-0047, 0065-0066, and 0076-0077] citations as in claim 1, above.).
Cai et al. and Yang et al. are considered to be analogous to the claimed invention because they are in the same field of endeavor in natural language data processing (e.g., entity/topic identification). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cai et al. to incorporate the teachings of Yang et al. of wherein the similarity score is calculated between the vectorized version of the candidate phrase and the vectorized version of the combined context data which provides the benefit of having superior performance for the task of keyphrase extraction ([0045] of Yang et al.).
Regarding claims 3 and 14, Cai et al. in combination with Yang et al. and Duboue et al. teaches the limitations as in claim 1 and 12, above.
Yang et al. further teaches:
3 and 14. The system/method of claims 1 and 12,
wherein the plurality of candidate phrases are extracted from the first field of the textual input using a graph ranking method (see ¶ [0039, 0048, and 0060]: “[0039] Unsupervised methods attempt to extract the underlying structure of the data without the assistance of previously labeled examples. Some unsupervised approaches that have been proposed include: (1) graph-based ranking methods, (2) topic-based clustering, (3) simultaneous learning, and (4) language modeling. For graph-based ranking methods, the importance of a candidate is determined by its “relatedness” to other candidates. Relatedness can be interpreted as co-occurrence or semantic relations (semantic-relatedness). A document may be represented by a network where the nodes are keyphrases. A candidate phrase is important if it is connected to important keyphrases or a large number of keyphrases. Then nodes are ranked based on their importance using a graph-based ranking method. Some examples of these algorithms include TextRank, DivRank, SingleRank, ExpandRank, CollabRank, among others. [0048] Some embodiments of the present disclosure include a keyphrase identification step, which may also be referred to as a candidate phrase identification step. In some embodiments, unsupervised, graph-based automatic keyword extraction models may be used to extract the most important words or terms from the input text. Models that may be used include YAKE (Yet Another Keyphrase Extraction method), Topical PageRank (TPR), TextRank, FirstPhrase, TF-IDF, and StupidKE. Each model may retrieve a ranked list of phrases and their scores. In other words, in embodiments, each model or any one or combination thereof may be utilized in sequence or in parallel. [0060] In some embodiments, one or more of unsupervised models 422 may be graph-based methods configured to, for example, build a graph based on input text such that nodes are words and edges represent a relation such as TF-IDF weight, co-occurrence, and/or position in text. The nodes may be ultimately ranked by their weight. For example, TextRank may be configured to build a graph with the words as nodes and edges representing co-occurrence relation, with the nodes ranked by their weight. TopicRank may be configured to build topics by creating HAC (average linkage), and then weighting the topics using random walk and selecting the first occurring candidate from important topics. Importance may be defined based on the PageRank algorithm. First Phrase or StupidKE may be configured to select sequences of nouns and adjectives and rank them according to inverse positions. YAKE may be configured to use statistical text features extracted from single documents to select words without linguistic tools or external resources.”).
Cai et al. and Yang et al. are considered to be analogous to the claimed invention because they are in the same field of endeavor in natural language data processing (e.g., entity/topic identification). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cai et al. to incorporate the teachings of Yang et al. of wherein the plurality of candidate phrases are extracted from the first field of the textual input using a graph ranking method which provides the benefit of having superior performance for the task of keyphrase extraction ([0045] of Yang et al.).
Regarding claims 4 and 15, Cai et al. in combination with Yang et al.and Duboue et al. teaches the limitations as in claim 1 and 12, above.
Cai et al. further teaches:
4 and 15. The system/method of claims 1 and 12,
wherein the textual input is a customer ticket (see ¶ [0008]: “…Specific technical approaches are described for automatically generating output data structures based on a received input data set representative of a new incident ticket…”) .
Regarding claims 5 and 16, Cai et al. in combination with Yang et al. teaches the limitations as in claims 4 and 15, above.
Cai et al. further teaches:
5 and 16. The system/method of claims 4 and 15,
wherein the first field is a title of the customer ticket (see ¶ [0011]: “The data receiver is configured for receiving the string inputs representative of the new incident ticket, the new incident ticket including at least a title field and a description field.”) .
Regarding claims 6 and 17, Cai et al. in combination with Yang et al.and Duboue et al. teaches the limitations as in claim 5 and 16, above.
Cai et al. further teaches:
6 and 17. The system/method of claims 5 and 16,
wherein the second field is a problem description field of the customer ticket (see ¶ [0011]: “The data receiver is configured for receiving the string inputs representative of the new incident ticket, the new incident ticket including at least a title field and a description field.”) .
Claim 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Cai et al. (US 20190347282 A1) and further in view of Yang et al. (US 20230091076 A1) and Duboue et al. (US 20120078902 A1) as applied to claim 12 above, and further in view of Bikel et all. (US 9324323 B1) and Moore (US 20040098247 A1).
Regarding claim 19, Cai et al. in combination with Yang et al. and Duboue et al. teaches the limitations as in claim 12, above.
Cai et al. further teaches:
19. The system/method of claim 12,
wherein the operations further comprise:
generating a third score for each potential candidate phrase based on how frequently words of the potential candidate phrase appear in the textual input and in the one or more data sources related to the textual input (see ¶ [0009 and 0011-0012, 0079, 0085-0086, 0090-0091, 0093, and 0164] citations as in claim 1 above. More specifically: ¶ [0085-0086]: “[0085] As the string corresponding to the new incident is likely not an exact match to any historical incident, both the new incident string and historical strings can be converted into vector representations. Term frequency inverse document frequency analysis is performed across against a set of historical incident-solution pairs stored on a data repository and the pre-processed concatenated title field and description field to establish an initial set of vectors for analysis. [0086] Term frequency inverse document frequency analysis is a weighted word process which helps in identifying word patterns (bigrams and trigrams) from a document/text. N-gram is a hyper parameter which is used in the term frequency inverse document frequency analysis approach to specify the word count range. N-gram range(1,3) specifies up to 3 word combination for the TFIDF. )
However, Cai et al. in combination with Yang et al. and Duboue et al. do not explicitly teach, but Bikel et al. does teach:
generating a first score for each potential candidate phrase based on the presence of technical terms in the potential candidate phrase (see ¶ Col. 11, lines 13-37: “(50) A speech recognition engine 322 processes the sounds collected from the user 304 to identify words included in the sounds and to provide phonetic descriptions of those words to a dictionary component, which provides actual text corresponding to the phonetic descriptions. The speech recognition engine 322 uses one or more of the topic language models to process the sounds to produce the textual data. For example, the topic identifier engine 320 may determine that the sounds are associated with the topic of “medicine”, and assign a relatively high topical relevance score to a topic model that includes medical terminology and jargon. The speech recognition engine 322 uses the statistical weighting to identify one or more of the topic language models 318a-318d for use in processing the sounds. Continuing the previous example, the speech recognition engine 322 may use the relatively high relevance score to identify and/or associate a relatively high statistical weight with a “medicine” topic language model, and use the “medicine” topic language model to recognize the speech such that recognition candidates for medical terms and jargon in the speech (e.g., anatomical terms, pharmaceutical names, pathogen names) may have a relatively higher likelihood of being selected as correctly identifying the words in the speech compared to recognition candidates corresponding to less relevant topics.”) ;
Cai et al.,Yang et al., and Duboue et al. and Bikel et al. are considered to be analogous to the claimed invention because they are in the same field of endeavor in natural language processing (e.g., topic/entity recognition/identification). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cai et al. in combination with Yang et al. and Duboue et al. to incorporate the teachings of Bikel et al. of generating a first score for each potential candidate phrase based on the presence of technical terms in the potential candidate phrase which provides the benefit of improving the quality of the output of a language model if the language model is built from training data that is about the same, or similar, topics (Col. 2, lines 46-48 of Bikel et al.).
However, Cai et al. in combination with Yang et al. Duboue et al., and Bikel et al. do not explicitly teach, but Moore does teach:
generating a second score for each potential candidate phrase based on if the potential candidate phrase is capitalized (see ¶ [0083 and 0107]: “[0083] Model 404 then optionally computes and applies a capitalization probability score which is added to the inside score and is a log probability estimated for several capitalization patterns. This is also discussed below in greater detail with respect to FIGS. 5A and 5B. Application of model 404 is illustrated by block 424 in FIG. 4. [0107] Model 404 can also apply an additional probability reflecting the likelihood of the capitalization pattern of the candidate phrase, if the phrase to be translated from the source language sentence is a captoid. This is, of course, optional, and can be eliminated where it is not desired. In any case, a number of different capitalization patterns are illustratively considered in this portion of model 404. For example, where the identified phrase is a captoid, and the candidate phrase has its first word capitalized, this is illustratively associated with a first, relatively high probability. Where none of the words in the candidate phrase are capitalized, this is illustratively associated with a second, lower probability. Finally, where the candidate phrase does not have its first word capitalized, but other words in the candidate phrase are capitalized, this is illustratively associated with a third, even lower probability. The capitalization probabilities are initially estimated from the candidate phrases that have the highest translation probability (highest inside score plus outside score) for each sentence pair and each identified source language phrase.”)
Cai et al., Yang et al., Duboue et al., Bikel et al., and Moore are considered to be analogous to the claimed invention because they are in the same field of endeavor in natural language processing (e.g., topic/entity recognition/identification). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cai et al. in combination with Yang et al., Duboue et al. and Bikel et al. to incorporate the teachings of Moore of generating a second score for each potential candidate phrase based on if the potential candidate phrase is capitalized which provides the benefit of enhances the derivation of translation relationships for captoids and other types of phrases ([0126] of Moore).
Allowable Subject Matter
Claims 7-11 and 18 objected to as being dependent upon a 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. As well as if rewritten to overcome any claim objections/rejections (i.e., 35 USC 101 abstract idea rejections) in this Office Action.
The following is a statement of reasons for the indication of allowable subject matter:
Regarding independent claims 7 and 18, the closest prior art of record Cai et al. (US 20190347282 A1), Yang et al. (US 20230091076 A1), and Duboue et al. (US 20120078902 A1) teach all of the limitations as in independent claims 1 and 12, above.
However, none of the cited Prior arts alone or in combination disclose the claim language as disclosed in claims 7 and 18:
7 and 18. The system/method of claims 1 and 12,
wherein extracting the plurality of candidate phrases from the first field of the textual input comprises:
extracting a plurality of potential candidate phrases from the first field of the textual input;
scoring each potential candidate phrase of the plurality of potential candidate phrases based at least on: a presence of technical terms in the potential candidate phrase, if the potential candidate phrase is capitalized, and a frequency of words of the potential candidate phrase in the first field of the textual input, the second field of the textual input, and one or more secondary data sources; and
selecting a subset of the plurality of potential candidate phrases having highest scores out of the plurality of potential candidate phrases.
(Emphasis added)
Hence, dependent claims 8-11 are also objected to as being dependent upon a 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. As well as if rewritten to overcome any claim objections/rejections (i.e., 35 USC 101 abstract idea rejections) in this Office Action.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Keisha Y Castillo-Torres whose telephone number is (571)272-3975. The examiner can normally be reached Monday - Friday, 9:00 am - 4:00 pm (EST).
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, Pierre-Louis Desir can be reached at (571)272-7799. 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.
Keisha Y. Castillo-Torres
Examiner
Art Unit 2659
/Keisha Y. Castillo-Torres/Examiner, Art Unit 2659
/PIERRE LOUIS DESIR/Supervisory Patent Examiner, Art Unit 2659
Prosecution Insights