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 .
DETAILED ACTION
This office action is in response to the claims filed on 03/21/2024.
Claims 1-20 are presented for examination.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35
U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 11, 12,13, 14, 18, 19, 20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA).
Regarding claim 1, IKEDA teaches an information processing apparatus comprising: readout processor circuitry configured to read out a tree- structured learned model that is obtained by performing a learning process on a tree- structured model by using a first data set(IKEDA, [Par.0028-0031], [Par.00052-0054], “[0028], The binary tree structure creation unit 11 creates a binary tree structure using a plurality of input data pieces. The score calculation unit 13 calculates a score using a node evaluation value for a node feature vector which is a feature of each node passing from a root node of the binary tree structure created by the binary tree structure creation unit 11 to a leaf node. The learning unit 14 learns a node evaluation model for calculating the node evaluation value for the node feature vector of each node of the binary tree structure” and [Par.0052-0054], “FIG. 6 shows an example in which the binary tree structure is created for the data pieces shown in FIG. 5. FIG. 6 shows a case in which 50% of data pieces shown in FIG. 5 (d1, d3, d5, d7, d9, and d11) is sampled to create the binary tree structure for the purpose of simplifying the explanation. The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied.[0053] A node (k=1) (k=1 is a node identifier) shown in FIG. 6 corresponds to the root node. The node (k=1) is branched using a feature f1 (POST rate). That is, under the condition “f1<0.5”, if this condition is satisfied, the data included in the node (k=1) is branched to an intermediate node (k=3), whereas if this condition is not satisfied, the data included in the node (k=1) is branched to the leaf node (k=2).” Examiner’s note, the tree structure is created based on the learning process on the input data, the learning unit is considered as the readout processor circuitry to determine the structure of the binary tree.) ;
leaf node identification processor circuitr(IKEDA, [Par.0052-0054], “FIG. 6 shows an example in which the binary tree structure is created for the data pieces shown in FIG. 5. FIG. 6 shows a case in which 50% of data pieces shown in FIG. 5 (d1, d3, d5, d7, d9, and d11) is sampled to create the binary tree structure for the purpose of simplifying the explanation. The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied.[0053] A node (k=1) (k=1 is a node identifier) shown in FIG. 6 corresponds to the root node. The node (k=1) is branched using a feature f1 (POST rate). That is, under the condition “f1<0.5”, if this condition is satisfied, the data included in the node (k=1) is branched to an intermediate node (k=3), whereas if this condition is not satisfied, the data included in the node (k=1) is branched to the leaf node (k=2).” And [par.0089-0090], “In this example embodiment, the node evaluation model learned by the learning unit 14 may be reused when the binary tree structure is reconstructed. That is, when there is a change in the data set due to an increase in data or the like, the binary tree structure is reconstructed as necessary, but at this time, the learned node evaluation model may be used.[0090] FIG. 12 is a flowchart for explaining the operation of the anomaly detection apparatus according to this example embodiment, and is a flowchart for explaining the operation when the binary tree structure is reconstructed. As shown in FIG. 12, when the binary tree structure is reconstructed, the binary tree structure creation unit 11 (see FIG. 2) of the anomaly detection apparatus 1 creates a binary tree structure using a data set for reconstruction (Step S11).” Examiner’s note, the node evaluation model is learn to identify the position of the branch node of the root node, whether the branch node (leaf node) will go to the right branch or the left branch in the tree structure, wherein the learned node evaluation model is reused to re-constructure the tree structure based on the changed in data (Second dataset) by the tree structure creation unit (leaf node identification processor circuitry));
and a ratio information generation generator configured to generate information related to a ratio between a number of all leaf nodes of the tree-structured learned model, and a number of the first leaf nodes or a number of second leaf nodes, the second leaf nodes being leaf nodes that do not each correspond to the first leaf node among the leaf nodes of the tree-structured learned model (IKEDA, [Fig.6, Par.0043], “Next, the score calculation unit 13 calculates a score y′ using the node evaluation value for the node feature vector which is the feature of each node passing from the root node to the leaf node of the binary tree structure (Step S2). For example, the node feature vector of each node is extracted using the node feature extraction unit 12. Further, for example, the node evaluation value is the weight for the node feature vector of each node, and the score calculation unit 13 calculates the score using the weight for the node feature vector of each node passing from the root node to the leaf node of the binary tree structure” Examiner’s note, the score calculation unit calculate the score based on the weight (feature vector ) of each node passing from the root node to the leaf node, therefore, the feature vector is considered as the generating information related to the ratio (score include the weight of each node passing from the root node to the all the leaf node of the tree structure). Fig.6 showing that the second leaf node(k=3) is not corresponding to the first leaf node( k=2) )).
Regarding claim 11 is rejected as the same reason as the claim 1, since these claims recite the same limitations.
Regarding claim 12 is rejected as the same reason as the claim 1, since these claims recite the same limitations.
Regarding claim 13 is rejected as the same reason as the claim 1, since these claims recite the same limitations.
Additionally, IKEDA further teaches a non-transitory computer readable storage medium encoded with computer readable instructions, which, when executed by processor circuitry, cause the processor circuitry to perform an information processing method according to the claim 12 (IKEDA, [Par.0098], “The program can be stored and provided to a computer using any type of non-transitory computer readable media. Non-transitory computer readable media include any type of tangible storage media. Examples of non-transitory computer readable media include magnetic storage media (such as floppy disks, magnetic tapes, hard disk drives, etc.), optical magnetic storage media (e.g. magneto-optical disks), CD-ROM (compact disc read only memory), CD-R (compact disc recordable), CD-R/W (compact disc rewritable), and semiconductor memories (such as mask ROM, PROM (programmable ROM), EPROM (erasable PROM), flash ROM, RAM (random access memory), etc.). The program may be provided to a computer using any type of transitory computer readable media. Examples of transitory computer readable media include electric signals, optical signals, and electromagnetic waves. Transitory computer readable media can provide the program to a computer via a wired communication line (e.g. electric wires, and optical fibers) or a wireless communication line.”).
Regarding claim 14, IKEDA teaches an information processing apparatus comprising: readout processor circuitry configured to read out a plurality of tree-structured learned models that is obtained by performing a learning process on a tree-structured model by using a first data setset(IKEDA, [Par.0028-0031, [Par.00052-0054], “[0028], The binary tree structure creation unit 11 creates a binary tree structure using a plurality of input data pieces. The score calculation unit 13 calculates a score using a node evaluation value for a node feature vector which is a feature of each node passing from a root node of the binary tree structure created by the binary tree structure creation unit 11 to a leaf node. The learning unit 14 learns a node evaluation model for calculating the node evaluation value for the node feature vector of each node of the binary tree structure” and [Par.0052-0054], “FIG. 6 shows an example in which the binary tree structure is created for the data pieces shown in FIG. 5. FIG. 6 shows a case in which 50% of data pieces shown in FIG. 5 (d1, d3, d5, d7, d9, and d11) is sampled to create the binary tree structure for the purpose of simplifying the explanation. The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied.[0053] A node (k=1) (k=1 is a node identifier) shown in FIG. 6 corresponds to the root node. The node (k=1) is branched using a feature f1 (POST rate). That is, under the condition “f1<0.5”, if this condition is satisfied, the data included in the node (k=1) is branched to an intermediate node (k=3), whereas if this condition is not satisfied, the data included in the node (k=1) is branched to the leaf node (k=2).” Examiner’s note, the tree structure is created based on the learning process on the input data, the learning unit is considered as the readout processor circuitry to determine the structure of the binary tree structure. The binary tree structure is corresponding to the plurality tree structures.) ;
leaf node identification processor circuitry configured to input a second data set to each of the tree-structured learned models and identify a first leaf node that is a leaf node corresponding to the second data set in each of the tree- structured learned models(IKEDA, [Par.0052-0054], “FIG. 6 shows an example in which the binary tree structure is created for the data pieces shown in FIG. 5. FIG. 6 shows a case in which 50% of data pieces shown in FIG. 5 (d1, d3, d5, d7, d9, and d11) is sampled to create the binary tree structure for the purpose of simplifying the explanation. The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied.[0053] A node (k=1) (k=1 is a node identifier) shown in FIG. 6 corresponds to the root node. The node (k=1) is branched using a feature f1 (POST rate). That is, under the condition “f1<0.5”, if this condition is satisfied, the data included in the node (k=1) is branched to an intermediate node (k=3), whereas if this condition is not satisfied, the data included in the node (k=1) is branched to the leaf node (k=2).” And [par.0089-0090], “In this example embodiment, the node evaluation model learned by the learning unit 14 may be reused when the binary tree structure is reconstructed. That is, when there is a change in the data set due to an increase in data or the like, the binary tree structure is reconstructed as necessary, but at this time, the learned node evaluation model may be used.[0090] FIG. 12 is a flowchart for explaining the operation of the anomaly detection apparatus according to this example embodiment, and is a flowchart for explaining the operation when the binary tree structure is reconstructed. As shown in FIG. 12, when the binary tree structure is reconstructed, the binary tree structure creation unit 11 (see FIG. 2) of the anomaly detection apparatus 1 creates a binary tree structure using a data set for reconstruction (Step S11).” Examiner’s note, the node evaluation model is learn to identify the position of the branch node of the root node, whether the branch node (leaf node) will go to the right branch or the left branch in the tree structure, wherein the learned node evaluation model is reused to re-constructure the tree structure based on the changed in data (Second dataset) by the tree structure creation unit);
and a ratio information generator configured to generate, for each of the tree-structured learned models, information related to a ratio between a number of all leaf nodes of each of the tree-structured learned models, and a number of the first leaf nodes or a number of second leaf nodes, the second leaf nodes being leaf nodes that do not each correspond to the first leaf node among the leaf nodes of the tree- structured learned model(IKEDA, [Par.0043], “Next, the score calculation unit 13 calculates a score y′ using the node evaluation value for the node feature vector which is the feature of each node passing from the root node to the leaf node of the binary tree structure (Step S2). For example, the node feature vector of each node is extracted using the node feature extraction unit 12. Further, for example, the node evaluation value is the weight for the node feature vector of each node, and the score calculation unit 13 calculates the score using the weight for the node feature vector of each node passing from the root node to the leaf node of the binary tree structure” Examiner’s note, the score calculation unit calculate the score based on the weight (feature vector ) of each node passing from the root node to the leaf node, therefore, the feature vector is considered as the generating information related to the ratio (score include the weight of each node passing from the root node to the all the leaf node of the tree structure) ). Fig.6 showing that the second leaf node(k=3) is not corresponding to the first leaf node( k=2));
Regarding claim 18 is rejected as the same reason as the claim 1, since these claims recite the same limitations.
Regarding claim 19 is rejected as the same reason as the claim 1, since these claims recite the same limitations.
Regarding claim 20 is rejected as the same reason as the claim 13, since these claims recite the same limitations.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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.
Claims 2, 4 are rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of DAIDO et al. (Pub. No. 20230032646-hereinafter, DAIDO).
Regarding claim 2, IKEDA teaches the information processing apparatus according to claim 1, further comprising output a branch condition corresponding to the second leaf node among the leaf nodes of the tree-structured learned mode, (IKEDA, [Par.0053-0056], “The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied. …The node (k=2) shown in FIG. 6 corresponds to the leaf node. Data reaching the node (k=2) among sample data pieces is a data piece of d11. Further, the minimum and maximum values of the feature f1 of the data piece reaching the node (k=2) among all the data pieces are 0.8 and 1.0, respectively. In FIG. 6, “f1 ∈ [0.8, 1.01]” is described. Similarly, the minimum and maximum values of the feature f2 of the data piece reaching the node (k=2) among all the data pieces are 5 and 100, respectively. In FIG. 6, “f2 ∈ [5, 1001]” is described..” Examiner’s note, if the feature vector condition of the data is satisfied at the condition of the K=3 (second leaf node), then the node is branched to the left with the solid line.).
However, IKEDA does not teach comprising condition output processor circuitry
On the other hand, DAIDO teaches condition output processor circuitry (DAIDO, [Par.0058], “The condition determination node setting reading unit 22 reads the condition determination node setting related to the condition determination node of the decision tree model to be used for inference, and outputs the condition determination node setting to the condition determination process unit 23. The condition determination node setting reading unit 22 initially reads the condition determination node setting related to the root node. Here, the “condition determination node setting” is setting information related to the condition determination executed in the condition determination node, and specifically includes a “feature amount”, a “condition determination threshold value”, and a “condition determination command”)..
IKEDA and DAIDO are analogous in arts because they have the same field of endeavor of generating the information based on the tree structure.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the output a branch condition corresponding to the second leaf node among the leaf nodes of the tree-structured learned mode, as taught by IKEDA, to include the comprising condition output processor circuitry configured to output a branch condition, as taught by DAIDO. The modification would have been obvious because one of the ordinary skills in art would be motivated to arrange the node in the tree structure based on the condition (DIADO, [Par.0059], “The condition determination process unit 23 acquires a feature amount included in the condition determination node setting acquired from the condition determination node setting reading unit 22 from the input data stored in the storage unit.”).
Regarding claim 4, IDEKA teaches the information processing apparatus according to claim 2, wherein the branch condition is a series of branch conditions from a root node to the second leaf node of the tree-structured learned model (IKEDA , [Par.0052-0054], “FIG. 6 shows an example in which the binary tree structure is created for the data pieces shown in FIG. 5. FIG. 6 shows a case in which 50% of data pieces shown in FIG. 5 (d1, d3, d5, d7, d9, and d11) is sampled to create the binary tree structure for the purpose of simplifying the explanation. The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied…. Minimum and maximum values of the feature f1 of the data reaching the node (k=1) among all the data pieces are 0.0 and 1.0, respectively. In FIG. 6, “f1 ∈ [0.0, 1.01]” is described. Similarly, minimum and maximum values of a feature f2 of the data reaching the node (k=1) among all the data are 2 and 140, respectively. In FIG. 6, “f2 ∈ [2, 140]” is described.” Examiner’s note, the Fig.6 shows the series of the condition from the root node to the leaf node in the binary tree structure.).
Claim 3, 8, 10 are rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of DAIDO et al. (Pub. No. 20230032646-hereinafter, DAIDO) and further in view of Timar et al. (Patent. No. 10783288 -hereinafter, Timar).
Regarding claim 3, IKEDA teaches the information processing apparatus according to claim 2, database for a data set that satisfies the branch condition corresponding to the second leaf node (IKEDA , [par,0052-0053], “FIG. 6 shows an example in which the binary tree structure is created for the data pieces shown in FIG. 5. FIG. 6 shows a case in which 50% of data pieces shown in FIG. 5 (d1, d3, d5, d7, d9, and d11) is sampled to create the binary tree structure for the purpose of simplifying the explanation. The threshold in each node is an average value of the minimum value and the maximum value of the data included in the node. In the binary tree structure shown in FIG. 6, solid arrows (branches to the left) show cases when conditions are satisfied, while broken arrows (branches to the right) show cases when conditions are not satisfied.[0053] A node (k=1) (k=1 is a node identifier) shown in FIG. 6 corresponds to the root node. The node (k=1) is branched using a feature f1 (POST rate). That is, under the condition “f1<0.5”, if this condition is satisfied, the data included in the node (k=1) is branched to an intermediate node (k=3), whereas if this condition is not satisfied, the data included in the node (k=1) is branched to the leaf node (k=2).” Examiner’s note, the root node identifies the whether the data include in the node (k=1) will be branched to the left node (second leaf node) or the right node based on the condition of each node.).
However, IKEDA does not teach further comprising data search processor circuitry configured to search a predetermined database
On the other hand, Timar teaches comprising data search processor circuitry configured to search a predetermined database (Timar, [Col.28, 44-65], The Web crawler 421 may search specific, known Web sites and may expand the list of Web sites based on results of information retrieved from the known Web sites. “),
IKEDA and Timar are analogous in arts because they have the same field of endeavor of generating the plurality of information.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to the information processing apparatus according to claim 2, database for a data set that satisfies the branch condition corresponding to the second leaf node, as taught by IKEDA, to include the data search processor circuitry configured to search a predetermined database, as taught by Timer. The modification would have been obvious because one of the ordinary skills in art would be motivated to search and gather the relevant information, (Timar, [col. 28, lines 47-59], “The search engine 420 may use command line interface 423 to search known databases. In addition to searching data and information sources, the subscription server 415 may receive push data and the streamer 422 may receive airport monitoring data in real time or near real time where such data is obtained from the airport of interest and is available to the PARC system 300. In block 513, the input module 342 identifies information and data relevant to, or possibly relevant to, a capacity estimation for the runway of interest, (i.e., the runway 22), the search engine 420 obtains the information and data and passes the information and data to the processing system 430 along with information and data from the subscription server 415.” ).
Regarding claim 8, IKEDA teaches , the information processing apparatus according to claim 1, the first data set, but it does not teach wherein the first data set is a learning data set and the second data set is an evaluation data set
On the other hand, Timar teaches wherein the first data set is a learning data set and the second data set is an evaluation data set (Timar, [Col.14, lines 45-48], “In an embodiment, the processed data are first divided into separate data sets, one for parameterizing the model, the other for verifying the model.” ).
IKEDA and Timar are analogous in arts because they have the same field of endeavor of generating the plurality of information.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to the first data set, as taught by IKEDA, to include the wherein the first data set is a learning data set and the second data set is an evaluation data set, as taught by Timar. The modification would have been obvious because one of the ordinary skills in art would be motivated to search and gather the relevant information, (Timar, [col. 28, lines 47-59], “The search engine 420 may use command line interface 423 to search known databases. In addition to searching data and information sources, the subscription server 415 may receive push data and the streamer 422 may receive airport monitoring data in real time or near real time where such data is obtained from the airport of interest and is available to the PARC system 300. In block 513, the input module 342 identifies information and data relevant to, or possibly relevant to, a capacity estimation for the runway of interest, (i.e., the runway 22), the search engine 420 obtains the information and data and passes the information and data to the processing system 430 along with information and data from the subscription server 415.” ).
Regarding claim 10, IKEDA teaches the information processing apparatus according to claim 1, the first data set, but it does not teach wherein the first data set and the second data set are derived from a same data set.
On the other hand, Timar teaches wherein the first data set and the second data set are derived from a same data set (Timar, [Col.14, lines 45-48], “In an embodiment, the processed data are first divided into separate data sets, one for parameterizing the model, the other for verifying the model.”).
IKEDA and Timar are analogous in arts because they have the same field of endeavor of generating the plurality of information.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to the first data set, as taught by IKEDA, to include the wherein the first data set and the second data set are derived from a same data set, as taught by Timar. The modification would have been obvious because one of the ordinary skills in art would be motivated to search and gather the relevant information, (Timar, [col. 28, lines 47-59], “The search engine 420 may use command line interface 423 to search known databases. In addition to searching data and information sources, the subscription server 415 may receive push data and the streamer 422 may receive airport monitoring data in real time or near real time where such data is obtained from the airport of interest and is available to the PARC system 300. In block 513, the input module 342 identifies information and data relevant to, or possibly relevant to, a capacity estimation for the runway of interest, (i.e., the runway 22), the search engine 420 obtains the information and data and passes the information and data to the processing system 430 along with information and data from the subscription server 415.” ).
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of DAIDO et al. (Pub. No. 20230032646-hereinafter, DAIDO) and further in view of Kalluri et al. (Pub. No. 20210263767-hereinafter, Kalluri).
Regarding claim 5, IKEDA teaches the information processing apparatus according to claim 1, but it does not teach comprising a count storage processor circuitry configured to store the number of times the second data set is associated for each of the leaf nodes of the tree-structured learned model.
On the other hand, DIADO teaches further comprising count storage processor circuitry configured to store the number of times the second data set (DAIDO, [Par.0070-0077], “Therefore, in the second example embodiment, the input data itself are stored in a storage unit or the like without being divided, while only the row numbers of the input data are collected to form a row number group, which is divided and passed to each child node. That is, each row number of the input data is used as a pointer to the input data stored in the storage unit, and pointers are grouped to perform the parallel process..”).
IKEDA and DAIDO are analogous in arts because they have the same field of endeavor of generating the information based on the tree structure.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing system, as taught by IKEDA, to include the count storage processor circuitry configured to store the number of times the second data set, as taught by DAIDO. The modification would have been obvious because one of the ordinary skills in art would be motivated to store the data in the data storage, (DAIDO, [Par.0070], “the condition determination process unit 23 compares the feature amount j for each data row of the input data Data by the function ‘compare’ with the condition determination threshold value (step S13-1). The data division unit 24 stores a data row regarded as a comparison result corresponding to a branch on a left side of the target node in divisional data LeftData in step S13-2, and stores a data row resulting in a comparison result corresponding to a branch on a right side of the target node in divisional data RightData in the step S13-3. The condition determination process unit 23 performs this process for all data rows of the input data Data, and terminates the loop process. This loop process is performed by the parallel process.”).
However, neither IKEDA nor DAIDO teach further comprising the number of times the second data set is associated for each of the leaf nodes of the tree-structured learned model,
On the other hand, Kalluri teaches further comprising the number of times the second data set is associated for each of the leaf nodes of the tree-structured learned model (Kalluri, [Par.0012, 0055], “[Par.0012], Implementations may include one or more of the following features. The computer-implemented method further including: determining a structure of the workflow, the structure being represented by a plurality of nodes of a tree structure, where two nodes of the plurality of nodes of the tree structure are connected by one or more stages, where each task of the one or more tasks of the workflow corresponds to a node of the plurality of nodes or a stage of the one or more stages; and[0055],The new communication workflow 310 may be inputted into workflow performance predictor 240 to generate a predicted performance value. The predicted performance value may represent the predicted task outcome of the new communication workflow 310, such as a percentage of users who perform a target action in response to receiving a communication triggered by a task included in the workflow, such as Tasks 3 or 5. Workflow performance predictor 240 may receive a composite feature vector of new communication workflow 310 generated by feature vector generator 220, and input the composite feature vector into a trained machine-learning model stored at machine-learning models 260.” Examiner’s note, calculating the percent of the user select the tasks, wherein, the task is represented by the node of the tree structure.
IKEDA, DAIDO and Kalluri are analogous in arts because they have the same field of endeavor of generating the information processing system.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the combined teaching of IKEDA and DAIDO of information processing apparatus, further comprising count storage processor circuitry configured to store the number of times the second data set, as set forth above, to include the number of times the second data set is associated for each of the leaf nodes of the tree-structured learned model, as taught by Kalluri. The modification would have been obvious because one of the ordinary skills in art would be motivated to determine the workflow structure, (Kalluri, [Par.0012], Implementations may include one or more of the following features. The computer-implemented method further including: determining a structure of the workflow, the structure being represented by a plurality of nodes of a tree structure, where two nodes of the plurality of nodes of the tree structure are connected by one or more stages, where each task of the one or more tasks of the workflow corresponds to a node of the plurality of nodes or a stage of the one or more stages.”).
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of DAIDO et al. (Pub. No. 20230032646-hereinafter, DAIDO) and further in view of Kalluri et al. (Pub. No. 20210263767-hereinafter, Kalluri) and further in view of JAGOTA et al. (Pub. No. 20180165281-hereinafter, JAGOTA).
Regarding claim 6, IKEDA teaches the information processing apparatus according claim 5, but it does not teach , further comprising second condition output processor circuitry configured to output a branch condition corresponding to a leaf node the number of times of which is smaller than or equal to a predetermined number of times among the first leaf nodes,
On the other hand, JAGOTA teaches further comprising second condition output processor circuitry configured to output a branch condition corresponding to a leaf node the number of times of which is smaller than or equal to a predetermined number of times among the first leaf nodes (JAGOTA, [Par.0041], “Associating the node in the trie with the record may include tokenizing the value stored in the field by the record; identifying each node, beginning from a root of the trie, corresponding to a token value sequence associated with the tokenized value, until a node is identified that stores a count less than a node threshold; identifying a branch sequence that includes each identified node as a key for the record; and associating the key with the node, and the record with the key. For example, the database system tokenizes the organization name National Institute of Health as <national, institute, of health> for a database record during the indexing phase. Continuing the example, the database system uses the tokenized values national, institute, of health to identify that a first sequential node in the trie 300 stores the count 3 for the token value sequence national, and stops after identifying that a second sequential node in the trie 300 stores the count 1 for the token value sequence national, institute, because this second sequential node's count 1 is less than the token threshold count of 2.5, as depicted in FIG. 3. Further to the example, the database system identifies the branch sequence national, institute in the trie 300 as a key for the database record that stores the organization name National Institute of Health. Concluding this example, the database system tags the node after the institute branch in the trie 300 with the key national institute, and adds the database record that stores the organization name National Institute of Health to a list of records for the key national institute.”).
IKEDA, DAIDO and JAGOTA are analogous in arts because they have the same field of endeavor of generating the information data.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing apparatus, as taught by IKEDA, to include the comprising second condition output processor circuitry configured to output a branch condition corresponding to a leaf node the number of times of which is smaller than or equal to a predetermined number of times among the first leaf nodes, as taught by JAGOTA. The modification would have been obvious because one of the ordinary skills in art would be motivated to generate the information on the tree structure based on the condition, (JAGOTA, [Par.0041], dentifying a branch sequence that includes each identified node as a key for the record; and associating the key with the node, and the record with the key. For example, the database system tokenizes the organization name National Institute of Health as <national, institute, of health> for a database record during the indexing phase. Continuing the example, the database system uses the tokenized values national, institute, of health to identify that a first sequential node in the trie 300 stores the count 3 for the token value sequence national, and stops after identifying that a second sequential node in the trie 300 stores the count 1 for the token value sequence national, institute, because this second sequential node's count 1 is less than the token threshold count of 2.5, as depicted in FIG. 3. Further to the example, the database system identifies the branch sequence national, institute in the trie 300 as a key for the database record that stores the organization name National Institute of Health. Concluding this example, the database system tags the node after the institute branch in the trie 300 with the key national institute, and adds the database record that stores the organization name National Institute of Health to a list of records for the key national institute.”).
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of Laaser et al. (Patent. No. 8280723-hereinafter, Laaser).
Regarding claim 7, IKEDA teaches the information processing apparatus according to claim 1, but it does not further comprising an error generator configured to generate, for each of the leaf nodes of the tree-structured learned model, an inference error between an output based on the leaf node and a ground truth value.
On the other hand, Laaser teaches further comprising an error generator configured to generate, for each of the leaf nodes of the tree-structured learned model, an inference error between an output based on the leaf node and a ground truth value (Laaser, [col.1, lines 43-59], “, for a given character associated with a second node in a second level in the tree structure, which is in a given branch of the tree structure that depends from the first level, the computer system: compares a second input character in the input string to the given character; generates a second label associated with the second node and the given branch, where the second label includes a second position in the input string, and a second cumulative error metric between the input string and the characters in the given branch; and prunes the given branch if the second cumulative error metric exceeds a predefined value. These operations of comparing, generating and pruning are repeated by the computer system for the remaining nodes in the second level in one or more branches. Moreover, the computer system continues to repeat the operations of comparing, generating and pruning for additional nodes in the one or more branches that depend from the nodes in the second level until a termination condition occurs.”).
IKEDA and Laaser are analogous in arts because they have the same field of endeavor of generating the information data.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing apparatus, as taught by IKEDA, to include the comprising an error generator configured to generate, for each of the leaf nodes of the tree-structured learned model, an inference error between an output based on the leaf node and a ground truth value, as taught by Laaser. The modification would have been obvious because one of the ordinary skills in art would be motivated to generate the information on the tree structure based on the error condition, (value (Laaser, [col.1, lines 43-59], “, for a given character associated with a second node in a second level in the tree structure, which is in a given branch of the tree structure that depends from the first level, the computer system: compares a second input character in the input string to the given character; generates a second label associated with the second node and the given branch, where the second label includes a second position in the input string, and a second cumulative error metric between the input string and the characters in the given branch; and prunes the given branch if the second cumulative error metric exceeds a predefined value. These operations of comparing, generating and pruning are repeated by the computer system for the remaining nodes in the second level in one or more branches. Moreover, the computer system continues to repeat the operations of comparing, generating and pruning for additional nodes in the one or more branches that depend from the nodes in the second level until a termination condition occurs.”).
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of Kikuchi et al. (Pub. No. 20200118027-hereinafter, Kikuchi).
Regarding claim 9, IKEDA teaches the information processing apparatus according to the claim 1, but it does not clarity that the wherein the first data set is an evaluation data set and the second data set is a learning data set,
On the other hand, Kikuchi teaches wherein the first data set is an evaluation data set and the second data set is a learning data set (Kikuchi, [Par.0049], “After the learning unit 132 completes learning of the first or second training data, the determination unit 133 determines, by using the first machine learning model in the machine learning model storage unit 124 and the evaluation data that is input from the first generating unit 131, whether the classification accuracy with respect to the evaluation data satisfies a desired level of accuracy. That is, the determination unit 133 evaluates the accuracy of cross-testing result obtained by using DT and determines whether the accuracy satisfies a desired level of accuracy.”)
IKEDA and Kikuchi are analogous in arts because they have the same field of endeavor of generating the information data.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing apparatus, as taught by IKEDA, to include the first data set is an evaluation data set and the second data set is a learning data set, as taught by Kikuchi. The modification would have been obvious because one of the ordinary skills in art would be motivated to have the two set of data, (Kikuchi, [Par.0049],” After the learning unit 132 completes learning of the first or second training data, the determination unit 133 determines, by using the first machine learning model in the machine learning model storage unit 124 and the evaluation data that is input from the first generating unit 131, whether the classification accuracy with respect to the evaluation data satisfies a desired level of accuracy. That is, the determination unit 133 evaluates the accuracy of cross-testing result obtained by using DT and determines whether the accuracy satisfies a desired level of accuracy..”).
Claims 15, 16, 17 are rejected under 35 U.S.C. 103 as being unpatentable over IKEDA et al. (Pub. No. US20220129764– hereinafter, IKEDA) in view of Phan et al. (Pub. No. 20220057786-hereinafter, Phan).
Regrading claim 15, IKEDA teaches the information processing apparatus according to claim 14, but it does not teach wherein the plurality of tree-structured learned models is obtained through ensemble learning.
On the other hand, Phan teaches wherein the plurality of tree-structured learned models is obtained through ensemble learning ([Par.0027], “A tree ensemble model combines predictions from multiple decision trees ƒ.sub.t(x). A decision tree uses a tree-like structure to predict the outcome for an input feature vector x. The t-th regression tree in the ensemble model has the following form”).
IKEDA and Phan are analogous in arts because they have the same field of endeavor of generating the information data based on the tree structure.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing apparatus, as taught by IKEDA, to include the plurality of tree-structured learned models is obtained through ensemble learning, as taught by Phan. The modification would have been obvious because one of the ordinary skills in art would be motivated to generate the tree structure model based on the ensemble learning, (Phan, [Par.0027], “A tree ensemble model combines predictions from multiple decision trees ƒ.sub.t(x). A decision tree uses a tree-like structure to predict the outcome for an input feature vector x. The t-th regression tree in the ensemble model has the following form”).
Regrading claim 16, IKEDA teaches the information processing apparatus according to claim 15, but it does not teach wherein the ensemble learning includes bagging learning or boosting learning.
On the other hand, Phan teaches wherein the ensemble learning includes bagging learning or boosting learning. (Phan, [Par.0003], “, individual decision trees often suffer from high variance predictions and can overfit the training data, which lead to a poor out-of-sample predictive accuracy, if there is no restriction in the size of tree. By using the bagging techniques, the tree ensemble regression function outputs predictions by taking the weighted sum of multiple decision trees as:.”).
IKEDA and Phan are analogous in arts because they have the same field of endeavor of generating the information data based on the tree structure.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing apparatus, as taught by IKEDA, to include the ensemble learning includes bagging learning or boosting learning, as taught by Phan. The modification would have been obvious because one of the ordinary skills in art would be motivated to generate the tree structure model based on the ensemble learning which include the bagging learning , (Phan, [Par.0003], “, individual decision trees often suffer from high variance predictions and can overfit the training data, which lead to a poor out-of-sample predictive accuracy, if there is no restriction in the size of tree. By using the bagging techniques, the tree ensemble regression function outputs predictions by taking the weighted sum of multiple decision trees as:”).
Regrading claim 17, IKEDA teaches the information processing apparatus according to claim 16, but it does not teach wherein the bagging learning includes a random forest.
On the other hand, Phan teaches wherein the bagging learning includes a random forest, (Phan, [Par.0003], “, individual decision trees often suffer from high variance predictions and can overfit the training data, which lead to a poor out-of-sample predictive accuracy, if there is no restriction in the size of tree. By using the bagging techniques, the tree ensemble regression function outputs predictions by taking the weighted sum of multiple decision trees as” and [Par.0035], “The control system 100 can solve for the set variable and flow variable based on applying a mixed-integer linear program, when the predictive models are piece-wise linear functions such as random forest, multivariate adaptive regression splines and fully connected feed-forward network.”).
IKEDA and Phan are analogous in arts because they have the same field of endeavor of generating information data based on tree structure.
Accordingly, it would have been obvious to one of the ordinary skills in the art before the effective filing date of the claimed invention to modify the information processing apparatus, as taught by IKEDA, to include the the bagging learning includes a random forest, as taught by Phan. The modification would have been obvious because one of the ordinary skills in art would be motivated to generate the tree structure model based on the apply the prediction model includes random forest , (Phan, [Par.0035], “the control system 100 can solve for the set variable and flow variable based on applying a mixed-integer linear program, when the predictive models are piece-wise linear functions such as random forest, multivariate adaptive regression splines and fully connected feed-forward network.”).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to EM N TRIEU whose telephone number is (571)272-5747. The examiner can normally be reached on Mon-Fri from 9:00-5:00.
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, Omar Fernandez Rivas can be reached on (571) 272-2589. 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.
/E.T./Examiner, Art Unit 2128
/OMAR F FERNANDEZ RIVAS/Supervisory Patent Examiner, Art Unit 2128
Prosecution Insights