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 .
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-9 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites “the selected prior queries.” There is no antecedent basis for “the selected prior queries.” This renders the claim vague and indefinite.
The same analysis applies to claims 4 and 7. The dependent claims are rejected because they inherit the deficiencies of their respective parent claims.
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.
Claim(s) 1-9 is/are rejected under 35 USC 103 as being unpatentable over Gutta US 12,141,181 in view of CN 116991869 (CN ‘869) and further in view of Ngan, Unlocking the Future of Data Analytics: How Query History Powers Text-to-SQL Transformation, Feb. 22, 2024.
With respect to claim 1, Gutta teaches “”(Currently Amended) A system comprising: a storage device; and a plurality of processing nodes, wherein at least one processing node of the plurality of processing nodes is configured to: receive a natural language query” in Fig. 5 items 510n; col. 5:22-37 (NL question 202 is NL query);
“deploy a [deep learning model ] wherein the vectorizing [deep learning model] is configured to” in col. 3;52-57, col. 5:22-37 and col. 12:46-60 (NL query vectorized by 1st LSTM (deep learning) model);
“convert the natural language query to a database language syntax representation using decoding constrained by a database language grammar derived from a target database schema” in col. 12:9-21 and col. 12:33-60 (“’how many analysts are in region 124’ and a corresponding training query ‘SELECT analyst FROM employee_table WHERE location=‘region 124’” teaches decoding NL to SQL; the decoding is constrained by SQL (database language grammar) derived from target database schema, i.e. “table schemas, primary keys, foreign keys or schema descriptions from a database based on the database identifier”);
“convert the natural language query to a vectorized natural language representation in the form of an embedding vector” in col. 3;52-57, col. 5:22-37 and col. 12:46-60 (NL query vectorized by 1st LSTM (deep learning) model);
“compare the vectorized natural language representation to vectorized query history information” in Col. 6:11-29: (Examiner finds the combination of output of the first question encoder (initial user query vectorization) and the second question encoder (vectorized history representation) strongly suggests comparing the two);
“and generate database syntax in response to the natural language query based on the comparison of the vectorized natural language representation to the vectorized query history representation information” in Fig. 2 202, 210, and 220; col. 6:11-29 (Combination of question vector and previous question vectors are used to generate database language query (database syntax));
“by conditioning the [model] on the selected prior queries as exemplars” in and in col. 12:9-33 (training questions and queries are used to examples (exemplars) to train the model; Examiner finds this teaches “conditioning”);
“execute the generated database syntax” Col. 4:28-41 (Database (structured) query (generated database syntax) used to return data).
It appears Gutta fails to explicitly teach “wherein the comparison is based on vector cluster analysis.”
However, CN-116991869- teaches “wherein the comparison is based on vector cluster analysis” on page 9
in the process of obtaining the target database table of the historical inquiry problem, projecting all the original data problems into the sentence vector sample space, using the clustering algorithm to divide the similar problems into the same cluster, and taking the mass centre of the cluster as the representative problem; for each input natural language question sentence, the natural language question sentence is mapped into sentence vector through the word embedding model, according to the distance between the sentence vector and the mass centre of the history problem, selecting the target database table corresponding to the history problem with the nearest distance, taking the database table ID as the target database table ID of the new input natural language question; for the natural language question sentence which cannot be matched successfully, using the learning model based on the adjacent semantic vector idea, training the historical problem and semantic vector, learning the mapping relation between the historical problem and the target database table, and predicting the target database table of the new input natural language question sentence, at the same time, continuously updating the model according to the relationship between the new input natural language question and the target database table, and improving the accuracy of prediction;
(Examiner finds “selecting the target database table corresponding to the history problem with the nearest distance” teaches comparing; Examiner finds “historical query questions” are previously entered user queries).
CN ‘869 and Gutta are analogous art because they are from the same field of endeavor as the claimed invention.
It would have been obvious to one skilled in the art before the effective filing date of the invention to modify the comparison in Gutta to include “wherein the comparison is based on vector cluster analysis” as taught” by CN ‘869.
The motivation would have been to improve query translation accuracy. See CN ‘869 pp. 9-10:
The method creatively uses the clustering algorithm to cluster the history problem and extract the representative problem, reduces the space dimension of the problem and reduces the calculation amount; at the same time, the optimal target database table is obtained by using the model based on the neighbour, which improves the accuracy and efficiency of the inquiry. Through the iteration training of the unmatched natural language question sentence, the invention can continuously improve the accuracy of the prediction model and avoid the dependence on the manual interference.
The beneficial effects of the above technical solution are as follows: in the process of obtaining the target database table of the historical query question based on the adjacent semantic vector idea, projecting all the original data question into the sentence vector sample space, then for each newly input natural language question sentence, after obtaining the natural language question sentence vector through coding, The sentence vector closest to the natural language question sentence vector and the database table ID are found in the sentence vector space, and the database table ID is used as the target database table ID of the newly input natural language question sentence. so as to find the target database table for the natural language question sentence through the adjacent thought.
It appears Gutta fails to explicitly teach an LLM.
However Ngan teaches an LLM on p. 1 first paragraph (Copilot, for example is an LLM).
Ngan and Gutta are analogous art because they are from the same field of endeavor as the claimed invention.
It would have been obvious to one skilled in the art before the effective filing date of the invention to modify the deep learning module in Gutta et al. to include an LLM.
The motivation would have been to “analyze and interpret large datasets effectively.” See Ngan p. 2
With respect to claim 2, CN ‘869 teaches “2. (New) The system of claim 1, wherein the vectorizing LLM is further configured to compare distances of vector clusters of vectorized query history information with the vectorized natural language representation”
on page 9
in the process of obtaining the target database table of the historical inquiry problem, projecting all the original data problems into the sentence vector sample space, using the clustering algorithm to divide the similar problems into the same cluster, and taking the mass centre of the cluster as the representative problem; for each input natural language question sentence, the natural language question sentence is mapped into sentence vector through the word embedding model, according to the distance between the sentence vector and the mass centre of the history problem, selecting the target database table corresponding to the history problem with the nearest distance,
The motivation to combine this element with claim 1 is the same as given in claim 1 above.
With respect to claim 3, Gutta teaches “3. (New) The system of claim 1, wherein the vectorizing LLM is further configured to convert the natural language query to a SQL representation” in Fig. 3 step 350 and col. 3:52-56.
Claims 4-6 and claims 7-9 are rejected for the reasons indicated above for claims 1-3.
Response to Arguments
With respect to patent eligibility, Examiner finds “compare the vectorized natural language representation to vectorized query history information wherein the comparison is based on vector cluster analysis” is an additional element that, when read as an ordered combination with the remaining elements of claim 1, reflect an improvement disclosed in at least paragraph 47 of the specification. This analysis applies to claim 4 and claim 7. As such, the 101 rejections are withdrawn because the claimed invention integrates the exception.
Applicant asserts the claimed invention is patentable over the prior art of record. Applicant’s assertion is not persuasive.
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 ALBERT M PHILLIPS, III whose telephone number is (571)270-3256. The examiner can normally be reached 10a-6:30pm EST M-F.
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, Ann J Lo can be reached at (571) 272-9767. 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.
Prosecution Insights