DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
In response to the amendment filed on May 4, 2026:
Claims 1, 3-4, 6-9, 11-12, and 14-20 are amended.
Claim 2 is canceled.
Claims 1, and 3-20 are pending.
Response to Arguments
In response to the remarks filed on May 4, 2026:
a. Objection to claim 11 is withdrawn.
b. Applicant’s remarks regarding the 35 U.S.C. 101 rejections of claims 1, 3-8, and 10-19 have been fully considered but are not persuasive.
Claims 1, 14, and 15 each is directed to the abstract idea of collecting and analyzing data to select a processing technique for performing a query. Each claim, as a whole, does not integrate that idea into a practical application and does not recite significantly more than the abstract idea itself. The recited determining a count of unique values, comparing that count to a threshold, and performing a query using a selected technique are high-level data analysis and decision making steps that amount to selecting how to process information based on a condition, which is a mental abstraction and a form of computer-implemented data processing.
Per step 2A – prong 1, the claim recites determining a count of unique values in a database column, comparing that count to a threshold, and selecting a query technique based on the comparison. Those limitations are directed to evaluating information and making a choice based on the evaluation, which can be characterized as an abstract idea of data analysis and conditional decision-making. The specification confirms that the invention concerns optimizing SQL queries using candidate sets and selecting a technique based on the count of unique values ([0003], [0007]-[0013], and [0070] of instant specification). However, the current claim language in each independent claim does not recite any specific unconventional data structure, specific query planner modification or other concrete technological mechanism for carrying out the selection.
Per step 2A – prong 2, the additional limitations in each claim do not integrate the abstract idea into a practical application. Each claim merely recites that the threshold is based on available computational power of the CPU but does not recite how the CPU’s computational power is measures, what specific hardware state is used, how the threshold is dynamically adjusted, or how the selected query technique changes the operation of the database system in a concrete, technical way. The recitation of a generic CPU and a database table is insufficient to show an improvement to the functioning of the computer itself or another technology. The specification’s discussion of improved processing time, reduced memory, and candidate set generation shows performance benefits from more detailed embodiments (i.e., claims 9, and 20). Each claim does not require a particular machine configuration, a new database architecture, or an improved query execution mechanism. Instead, each claim merely uses a computer to carry out the abstract idea of selecting one query technique over another based on data count and a threshold.
Per step 2B, each claim also fails to recite significantly more than the abstract idea. The additional element s are not more than conventional computer components and routing database query operations (i.e., a CPU, a database table, a count of unique values, and a selected query technique). These limitations are implemented a high level of generality and do not amount to an inventive concept. Since the ordered combination of the actual claim limitations remains a generic instruction to perform a data comparison and use the comparison to choose a query path, each claim is insufficient to transform the abstract idea into patent-eligible subject matter.
Rejections of claims 3-8, 10-13, and 16-19 are maintained for the similar reasons presented previously and in view of their respective independent claims.
Rejections of claims 9, and 20 are withdrawn.
c. Applicant’s remarks regarding the 35 U.S.C. 102(a)(1) and 103 rejections of claims 1, and 3-20 have been fully considered but are moot in view of a new ground of rejections presented here on.
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.
The claimed invention in claims 1, 3-8, and 10-19 are directed to a judicial exception (i.e., an abstract idea) without significantly more.
Claims 1, 3-8, and 10-19 pass step 1 of the 35 U.S.C. 101 analysis since each claim is either directed to a method, non-transitory computer readable medium, or an apparatus comprising a memory and at least one processor (i.e., hardware components [00107] of instant specification).
Claims 1, 14, and 15 each is directed to the abstract idea of collecting and analyzing data to select a processing technique for performing a query. Each claim, as a whole, does not integrate that idea into a practical application and does not recite significantly more than the abstract idea itself. The recited determining a count of unique values, comparing that count to a threshold, and performing a query using a selected technique are high-level data analysis and decision making steps that amount to selecting how to process information based on a condition, which is a mental abstraction and a form of computer-implemented data processing.
Per step 2A – prong 1, the claim recites determining a count of unique values in a database column, comparing that count to a threshold, and selecting a query technique based on the comparison. Those limitations are directed to evaluating information and making a choice based on the evaluation, which can be characterized as an abstract idea of data analysis and conditional decision-making. The specification confirms that the invention concerns optimizing SQL queries using candidate sets and selecting a technique based on the count of unique values ([0003], [0007]-[0013], and [0070] of instant specification). However, the current claim language in each independent claim does not recite any specific unconventional data structure, specific query planner modification or other concrete technological mechanism for carrying out the selection.
Per step 2A – prong 2, the additional limitations in each claim do not integrate the abstract idea into a practical application. Each claim merely recites that the threshold is based on available computational power of the CPU but does not recite how the CPU’s computational power is measures, what specific hardware state is used, how the threshold is dynamically adjusted, or how the selected query technique changes the operation of the database system in a concrete, technical way. The recitation of a generic CPU and a database table is insufficient to show an improvement to the functioning of the computer itself or another technology. The specification’s discussion of improved processing time, reduced memory, and candidate set generation shows performance benefits from more detailed embodiments (i.e., claims 9, and 20). Each claim does not require a particular machine configuration, a new database architecture, or an improved query execution mechanism. Instead, each claim merely uses a computer to carry out the abstract idea of selecting one query technique over another based on data count and a threshold.
Per step 2B, each claim also fails to recite significantly more than the abstract idea. The additional element s are not more than conventional computer components and routing database query operations (i.e., a CPU, a database table, a count of unique values, and a selected query technique). These limitations are implemented a high level of generality and do not amount to an inventive concept. Since the ordered combination of the actual claim limitations remains a generic instruction to perform a data comparison and use the comparison to choose a query path, each claim is insufficient to transform the abstract idea into patent-eligible subject matter.
Claims 3, and 16 recite in each claim additional elements of accessing a vector corresponding to a predicate of the query; selecting a row of the database table; accessing an entry in the selected row…; accessing a stored vector…; computing…a semantic similarity…; and storing a result of the computer semantic similarity… which are all implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., visually viewing a vector drawn on paper of a predicate of the query; mentally selecting a row of the database table, visually viewing an entry in the selected row; visually view a vector drawn on paper corresponding to the viewed entry; mentally computing the semantic similarity between the vectors; and writing down result of the computed similarity on paper). Thus, the claims are ineligible. Even assuming that the stored vector and storing a result of the computed semantic similarity… are involved a physical storage, these elements involving physical storage are considered as extra-solution activities (per step 2A – prong 2 of the Abstract Idea Analysis) that cannot be integrated into a practical application (e.g., the elements recite trivial elements that occurred or would occur after the mental process) since such limitation(s) is/are no more than mere instructions to apply the exception using a generic computer component (e.g., processor, memory, and computer-executable instructions). The extra-solution activities in step 2A - prong 2 are reevaluated in step 2B to determining if each limitation is more than what is well-understood, routine, conventional activity in the field. The background of the limitations does not provide any indication that the computer components (e.g., processor, memory, and computer-executable instructions) are not off-the-shelf computer components. The Symantec, TLI, and OOP Techs court decisions cited in MPEP 2106.05(d)(II) indicate that mere receiving, generating, storing, determining, identifying, and transmitting of data over a network are a well-understood, routine, and conventional functions when claimed in a merely generic manner (as it is here). Accordingly, a conclusion that the claims are well-understood, routine, conventional activity is supported under Berkheimer Option 2. For these reasons, there is no inventive concept in each claims, thus, the claims are ineligible.
Claim 4 recites an additional element …accessing the stored result in response to the entry…being a repeated occurrence… which is implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., visually viewing the computed result written on paper based on an occurrence condition when performing the mental search). Thus, the claim is ineligible.
Claim 5 merely defines a range of the result corresponding to a range of similarity for the mental search analyzed above. Thus, the claim is ineligible.
Claims 6, and 17 recite in each claim additional elements of pre-calculating a plurality of semantic similarities by: accessing…a vector corresponding to each unique value of the unique pair…; computing…a semantic similarity between the two…vectors; and storing…a result of the computed semantic similarity which are all implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., visually viewing a vector drawn on paper corresponding to each unique value of the unique pair; mentally computing a semantic similarity between the viewed vectors for each unique pair; and writing down result of the computed semantic similarity on paper). Thus, the claims are ineligible.
Claims 7, and 18 recite in each claim additional elements of accessing a predicate of the query; selecting a row of the database table; accessing an entry in the selected row…; accessing the stored result… which are all implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., visually viewing a predicate of the query, mentally selecting a row of the database table; visually viewing an entry in the selected row; visually viewing the stored result). Thus, the claims are ineligible.
Claims 8, and 19 recite in each claim additional elements of accessing an entry in each row…; accessing a stored vector corresponding to a specific value…; clustering rows of the database table together…; storing a cluster identifier and a corresponding centroid value… which are all implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., visually viewing an entry in each row; visually viewing a stored vector corresponding to a specific value of each entry; mentally grouping the rows of the database table together; and writing down on paper a cluster ID and a representative value of the cluster for each candidate set). Thus, the claims are ineligible.
Claim 10 merely recites the clustering is k-means clustering specifying a k-means algorithm to be utilized for the mental search analyzed above. Thus, the claim is ineligible.
Claim 11 merely recites the computation of the semantic similarity is implemented with a dot product calculation specifying a dot product to be utilized for the mental computation analyzed above. Thus, the claim is ineligible.
Claim 12 recites additional elements of converting each unique entry…to a corresponding vector; and storing each corresponding vector in a vector table indexed by a value…. which are implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., mentally converting each entry in to a vector; and writing down each vector in a table format on table with a certain indexed value). Thus, the claim is ineligible.
Claim 13 recites an additional element of performing the query further comprises processing rows of a candidate set in batches…. which is implementable in a human mind and/or with the aid of pen/paper similar to the above analysis (e.g., mentally processing multiple rows of a candidate sets to perform the mental search). Thus, the claim is ineligible.
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.
Claims 1, 14, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Bestgen et al. (Pub. No. US 2009/0094192, published on April 9, 2009; hereinafter Bestgen) in view of Kadiam et al. (Pub. No. US 2020/0285643, published on September 10, 2020; hereinafter Kadiam).
Regarding claims 1, 14, and 15, Bestgen clearly shows and discloses a method (Abstract); a non-transitory computer readable medium comprising computer executable instructions which when executed by a computer cause the computer to perform the method; and an apparatus comprising: a memory; and at least one processor, coupled to said memory, and operative to perform operations (Figure 2) comprising:
determining, by a central processing unit (CPU), a count of unique values in a column of a database table (Optimizer 254 optimizes the execution of SQL queries against DBMS 250. Optimizer 254 is implemented as computer program instructions that optimize execution of a SQL query in dependence upon database management statistics 264. Database statistics are typically implemented as metadata of a table, such as, for example, metadata of tables of database 262 or metadata of database indexes. Database statistics may include, for example: cardinality statistics: a count of the number of different values in a column, [0048]-[0051]);
performing, by the CPU, a query on the database table, wherein a technique for performing the query is selected based on the count of unique values (The adaptive query processing module 150 includes computer program instructions capable of identifying poorly performing queries; substituting an alternate plan to execute the query; and executing the query using the alternate plan, [0058]. It is clear that the alternate plan is selected based at least on determined cardinality statistics).
Kadiam then discloses:
comparing, by the CPU, the count of unique values with a specific threshold, wherein the specific threshold is based on available computational power of the CPU (in determining the cost at action 212, optionally at action 218, the cost of the physical operators can be determined. In an example, cost computing component 126, e.g., in conjunction with processor 104, memory 106, database system 108, query optimizer 112, etc., can determine the cost of the physical operators. For example, cost computing component 126 can determine a cost for a set of the physical operators based on the cardinality parameters, a function result, available system resources (e.g., processor, memory, input/output, etc., [0076]); and
performing, by the CPU, a query on the database table, wherein a technique for performing the query is selected based on the comparison of the count of unique values with the specific threshold (execute alternative sub-plans according to actual step results determined at runtime, to adapt to the memory or performance needs of the query depending on cardinality estimates computed at run time, [0091]).
It would have been obvious to an ordinary person skilled in the art at the time of the invention was effectively filed to incorporate the teachings of Kadiam with the teachings of Bestgen for the purpose of establishing the query technique selection threshold based on the available computational power to ensure that high-overhead and/or resource-intensive execution branches are filtered to prevent query execution failures under heavy workloads.
Claims 3, 5, 8, 10-11, 13, 16, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Bestgen in view of Kadiam and further in view of Bandyopadhyay et al. (Pub. No. US 2018/0267977, published on September 20, 2018; hereinafter Bandyopadhyay).
Regarding claims 3, and 16, Bandyopadhyay then discloses:
accessing, by the CPU, a vector corresponding to a predicate of the query (Some embodiments enable CI queries using the word vectors in the vector space as user-defined functions (UDFs), [0027]. A query to identify similar customers would examine the word vectors for each customer (i.e. custA, custB, custC, custD), [0030]-[0031]);
selecting, by the CPU, a row of the database table (To prepare the WFFD for CI queries, the nutrients were partitioned into groups (vitamins, amino acids, etc.). The numeric values were grouped into clusters using K-means, and the word2Vec model was trained using 200 dimensions, [0049]);
accessing, by the CPU, an entry in the selected row (A query to identify similar customers would examine the word vectors for each customer (i.e. custA, custB, custC, custD). So, for custD, the relevant row (tuple) 404 would be “custD 9/16 Walmart NY Stationery ‘Crayons, Folders’ 25”, [0030]), the entry corresponding to a column identified by the query (The columns contain information such as ingredients, categories, nutrients, etc., [0048]. Similarity queries were run over ingredients (text), nutrients (text) and country (text), [0049]);
accessing, by the CPU, a stored vector corresponding to a specific value of the accessed entry in the selected row (In the vector space, the word vector of custD is more similar to the word vector of custB as both bought stationery, including crayons. Likewise, the word vector of custA is more similar to the word vector of custC as both bought fresh produce, including bananas, [0030]);
computing, by the CPU, using cosine similarity (When comparing two sets of vectors, similarity UDFs may be used to output a scalar similarity value. Similarity measures between any pair of vectors are determined using cosine and max-norm algorithms, [0045]), a semantic similarity between the vector corresponding to the predicate of the query and the stored vector corresponding to the entry in the selected row in response to the entry in the selected row being a first occurrence of encountering the specific value during the performance of the query (a query to identify similar customers would determine that custA is more similar to custD as both purchased goods in NY on 9/16 for similar amounts. Likewise, custB is now more similar to custC as both purchased goods on 10/16 for similar amounts, [0030]-[0031]); and
storing, by the CPU, a result of the computed semantic similarity in response to computing the semantic similarity (Results for products having similar ingredients and similar nutrients in similar countries is shown in Table 2. For example, Special K original from Kellogg's is similar to Crispy Flakes with Red Berries Cereal from Market Pantry in the United States, [0051]-[0052]).
It would have been obvious to an ordinary person skilled in the art at the time of the invention was effectively filed to incorporate the teachings of Bandyopadhyay with the teachings of Bestgen, as modified by Kadiam, for the purpose of adapting a relational database containing multiple data types to enhance processing of a query based on relationship amongst a set of representative vectors such that relevant results are returned corresponding to the query.
Regarding claim 5, Bandyopadhyay further discloses the result varies between 1.0 and -1.0, with 1.0 representing a greatest similarity and -1.0 representing a smallest similarity (When comparing two sets of vectors, similarity UDFs may be used to output a scalar similarity value. Similarity measures between any pair of vectors are determined using cosine, [0045]. It is well-known that the range of similarity scores calculated using cosine similarity (i.e., cosine of the angle between two vectors) is typically range from -1 to 1 with 1 being maximum similarity and -1 being maximum dissimilarity).
Regarding claims 8, and 19, Bandyopadhyay then discloses:
accessing, by the CPU, an entry in each row of the database table, each entry corresponding to a given column (A query to identify similar customers would examine the word vectors for each customer (i.e. custA, custB, custC, custD), [0030]-[0031]);
accessing, by the CPU, a stored vector corresponding to a specific value of the accessed entry corresponding to each row of the database table (So, for custD, the relevant row (tuple) 404 would be “custD 9/16 Walmart NY Stationery ‘Crayons, Folders’ 25”. In the vector space, the word vector of custD is more similar to the word vector of custB as both bought stationery, including crayons, [0030]-[0031]);
clustering, by the CPU, rows of the database table together into a plurality of candidate sets based on a semantic similarity of the accessed vector corresponding to each row of the database table (To prepare the WFFD for CI queries, the nutrients were partitioned into groups (vitamins, amino acids, etc.). The numeric values were grouped into clusters using K-means, and the word2Vec model was trained using 200 dimensions. Similarity queries were run over ingredients (text), nutrients (text) and country (text). Both the single-model and the ensemble approach were used, [0049]); and
storing, by the CPU, a cluster identifier and a corresponding centroid value for each of the candidate sets (For example, in a relational database having a number representing sales dollars, the actual dollar amount is converted to a cluster ID and expressed as “sales_clusterlD”. So, an actual token value of 5000 may be expressed as “sales_272” where 272 is the cluster ID of the cluster containing 5000, [0038].
Regarding claim 10, Bandyopadhyay further discloses the clustering is k-means clustering (Any traditional clustering algorithm may be used to cluster data (e.g., K-means, hierarchical clustering, etc.), [0038]).
Regarding claim 11, Bandyopadhyay then discloses the computation of the semantic similarity is implemented with a dot product calculation (When comparing two sets of vectors, similarity UDFs may be used to output a scalar similarity value. Similarity measures between any pair of vectors are determined using cosine, [0045]. It is well-known that, by definition, the cosine similarity is the dot product of two vectors divided by the products of their magnitudes).
Regarding claim 13, Bandyopadhyay further discloses the performing the query further comprises processing rows of a candidate set in batches (For the ensemble approach, more than one embedding model or clustering strategy (discussed below) are used for different data types (e.g., latitude/longitude, images or time-series). A default clustering approach or user-provided similarity functions may be used. The results are computed for each model or clustering group and final results are computed by merging multiple result sets. The final results are merged by finding the intersection between row-sets that represent results for each clustering group, [0034]).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Bestgen in view of Kadiam and further in view of Bandyopadhyay and further in view of Hu et al. (Pub. No. US 2021/0248258, published on August 21, 2021; hereinafter Hu).
Regarding claim 4, Hu then discloses accessing, by the CPU, the stored result in response to the entry in the selected row being a repeated occurrence of encountering the specific value during the performance of the query (Historical access result data may be retrieved from the data structure for each of the subset of detectors used for the current access request. Historical access result data may include a history of outcomes (e.g., accepted or rejected) for past access requests and a plurality of data elements. In some embodiments, the one or more data elements associated with the current access request may not be the same or only a subset of the plurality of data elements within the historical access result data. Furthermore, the historical access result data may be stored within a vector or in any other suitable manner, [0134]).
It would have been obvious to an ordinary person skilled in the art at the time of the invention was effectively filed to incorporate the teachings of Kadiam with the teachings of Bestgen, as modified by Bandyopadhyay and Hu, for the purpose of utilizing a stored vector of historical result values to determine whether a past result can be reused based on matching attributes associated with a current query.
Claims 6, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Bestgen in view of Kadiam and further in view of Douglas (Pub. No. US 2018/0374563, published on December 27, 2018).
Regarding claims 6, and 17, Douglas then discloses comprising pre-calculating a plurality of semantic similarities by:
accessing, by the CPU, for each unique pair of unique values from a column of the database table, a vector corresponding to each unique value of the unique pair (facilitating record matching and entity resolution and for enabling improvements in record linkage including determining records that refer to the same entity or individual as one or more other records in a collection of records that are stored in a computer system and detecting matches of a new record with one or more others that already exist and are stored in online databases. In an embodiment, a phenotypic bit-vector “fingerprint” pattern-specific weight is incorporated into conventional record linkage methods to enhance the record linkage accuracy and statistical performance, [0012], [0029]), the column being identified by the query (for each entity, a record linkage weight (rl_wt), shown in column 307, is determined for each row. Column 310 shows a combined composite weight of the rl_wt and ps_wt. In this example embodiment, RMS is used to determine the composite weight or score. Furthermore, here the scores are normalized to (0,1), [0072]-[0073]);
computing, by the CPU, for each unique pair of unique values, a semantic similarity between the accessed vectors corresponding to each unique value of the unique pair (Binary fingerprints are formed by (a) constructing bit-vectors (“fingerprints”) by calculating similarities or distances for each such combination and combining each pairwise similarity or distance with the corresponding conventional record-linkage weights, such as by using a root-mean-square, dot product cosine measure, [0014]. Outputting the unique identifiers of record matches identified by the pair-wise matching algorithm, [0088]); and
storing, by the CPU, for each unique pair of unique values, a result of the computed semantic similarity (modified Tanimoto similarity determination from steps associate with 290, such as by fingerprint calculation, may be treated as one marker, indicator, or ‘weight’ that measures the similarity of a record associated with the current entity to records from putative matching entities stored in the target database, [0067]).
It would have been obvious to an ordinary person skilled in the art at the time of the invention was effectively filed to incorporate the teachings of Kadiam with the teachings of Bestgen, as modified by Douglas, for the purpose of facilitating record matching for enabling improvements in record linkage including determining records that refer to the same entity as one or more other records in a collection of records that are stored in a computer system and detecting matches of a new record with one or more others that already exist and are stored in different databases.
Claims 7, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Bestgen in view of Kadiam in view of Douglas and further in view of Bandyopadhyay.
Regarding claims 7, and 18, Bandyopadhyay then discloses:
accessing, by the CPU, a predicate of the query (Some embodiments enable CI queries using the word vectors in the vector space as user-defined functions (UDFs), [0027]. A query to identify similar customers would examine the word vectors for each customer (i.e. custA, custB, custC, custD), [0030]-[0031]);
selecting, by the CPU, a row of the database table (To prepare the WFFD for CI queries, the nutrients were partitioned into groups (vitamins, amino acids, etc.). The numeric values were grouped into clusters using K-means, and the word2Vec model was trained using 200 dimensions, [0049]);
accessing, by the CPU, an entry in the selected row (A query to identify similar customers would examine the word vectors for each customer (i.e. custA, custB, custC, custD). So, for custD, the relevant row (tuple) 404 would be "custD 9/16 Walmart NY Stationery 'Crayons, Folders' 25", [0030]), the entry corresponding to a column identified by the query (The columns contain information such as ingredients, categories, nutrients, etc., [0048]. Similarity queries were run over ingredients (text), nutrients (text) and country (text), [0049]); and
accessing, by the CPU, the stored result corresponding to the unique pair that includes both the predicate and the entry (In the vector space, the word vector of custD is more similar to the word vector of custB as both bought stationery, including crayons. Likewise, the word vector of custA is more similar to the word vector of custC as both bought fresh produce, including bananas, [0030]. When comparing two sets of vectors, similarity UDFs may be used to output a scalar similarity value. Similarity measures between any pair of vectors are determined using cosine and max-norm algorithms, [0045]. CI queries may be used in a number of retail cases, such as customer analytics to find similar customers based on buying patterns (e.g., purchased items, frequency, amount spent, etc.). CI queries may also be used for advanced sales predictions using external data to predict sales of a new item being introduced based on sales of related or similar items currently being sold. CI queries may also be used to analyze historical sales data using external data, [0051]-[0052]).
It would have been obvious to an ordinary person skilled in the art at the time of the invention was effectively filed to incorporate the teachings of Bandyopadhyay with the teachings of Bestgen, as modified by Kadiam and Douglas, for the purpose of adapting a relational database containing multiple data types to enhance processing of a query based on relationship amongst a set of representative vectors such that relevant results are returned corresponding to the query.
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Bestgen in view of Kadiam and further in view of Cason et al. (Pub. No. US 2020/0410050, published on January 27, 2019; hereinafter Cason).
Regarding claim 12, Cason then discloses: converting, by the CPU, each unique entry in the database table to a corresponding vector; and storing , by the CPU, the corresponding vector of each unique entry in a vector table indexed by a value of the unique entry (vector table 617 includes vectors 618A and 618B. Vector 618A is the result of one-hot encoding of row 608I in consideration table 614, and vector 618B is the result of one-hot encoding of row 608J in consideration table 614. Vector table 617 has been constructed as though the traversals that produced rows 608I and 608J are the only traversals present in the training data set. Thereby, the columns 610 represent every unique value present in consideration table 614 (i.e., every combination of an attribute type and the attribute value is represented, although not all of the columns 610 are visible in FIG. 6C), [0162]).
It would have been obvious to an ordinary person skilled in the art at the time of the invention was effectively filed to incorporate the teachings of Kadiam with the teachings of Bestgen, as modified by Cason, for the purpose of converting the training data into training conversion tables and using one-hot encoding the training conversion tables to generate training vectors to accurately applying classification to data within a query pipeline.
Allowable Subject Matter
Claims 9, and 20 would be allowable if rewritten to incorporate the limitations of the respective independent claims and all intervening claims.
Relevant Prior Art
The following prior art is/are deemed relevant to the claims:
Al-Omari et al. (Pub. No. US 2012/0117055) teaches categorizing, via a computing system having a plurality of interconnected central processing units (CPUs) and a memory storing a join input data set, the join input data set into a high-skew data set and a low-skew data set during a query optimization phase by comparing an occurrence frequency of a join column value to a threshold value, wherein computing the threshold value comprises dividing an operator cardinality by a number of CPUs that the data will be distributed to; at query execution time, distributing the low-skew data set to the plurality of CPUs using a first distribution method.
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.
Contact Information
Any inquiry concerning this communication or earlier communications from the Examiner should be directed to Son Hoang whose telephone number is (571) 270-1752. The Examiner can normally be reached on Monday – Friday (7:00 AM – 4:00 PM).
If attempts to reach the Examiner by telephone are unsuccessful, the Examiner’s supervisor, Sherief Badawi can be reached on (571) 272-9782. 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 the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/SON T HOANG/
Primary Examiner, Art Unit 2169 June 18, 2026
Prosecution Insights