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 .
Claims 1-20 are pending.
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 (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 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)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(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.
Claim(s) 1-2, 4, 6-9, 11, 13-17, 19-20 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Bhadouria et al (US 20170300862 A1).
Regarding claim 1, Bhadouria substantially discloses, teaches or suggests a method for generating custom industry classification, comprising:
receiving, by an industry classification computer program executed by an electronic device, standard industry classifications and standard industry descriptions (see at least Fig.4 block 406);
receiving, by the industry classification computer program, custom industry classifications, custom industry descriptions (see at least Fig.4 block 414), and
a mapping of the custom industry classifications to the standard industry classifications, wherein the mapping maps a first portion of the custom industry classifications to the standard industry classifications (see at least Fig.5 blocks 504, 506, 508, 510);
generating, by the industry classification computer program, a dataset comprising standard industry descriptions as inputs and custom industry classifications as outputs (see at least Fig.5 blocks 512, 514, 516, 518, 520);
converting, by the industry classification computer program, the standard industry descriptions and unmapped portions of the custom industry descriptions to vector representations (see at least Fig.5 blocks 522, 524, 526, 528);
performing, by the industry classification computer program, similarity matching between the vector representations of the standard industry descriptions and the vector representations of the unmapped portions of the custom industry descriptions (see at least [0038]: The machine learning algorithm 412 may be selected from among many different potential supervised or unsupervised machine learning algorithms. Examples of supervised learning algorithms include artificial neural networks, Bayesian networks, instance-based learning, support vector machines, random forests, linear classifiers, quadratic classifiers, k-nearest neighbor, decision trees, and hidden Markov models, Examples of unsupervised learning algorithms include expectation-maximization algorithms, vector quantization, and information bottleneck method. In an example embodiment, a multi-class logistical regression model is used.);
assigning, by the industry classification computer program, each of the unmapped portions of the custom industry descriptions to one of the standard industry descriptions based on the similarity matching (see at least [0047] At operation 610, the TF-IDF values are L2 normalized to convert them to unit vectors. At operation 612, the set of features and their associated normalized TF-IDF values are output. As will be seen, the TD-IDF values will be later used during the creation of a feature index IDF mapping file that is used to create the IT-IDF features from the query terms for the job posting under test.);
training, by the industry classification computer program, a supervised classifier using the dataset and a plurality of startup company descriptions for a plurality of startup companies as inputs (see at least [0048]: Referring back to FIG. 5, at operation 506, data partitioning is performed. This process partitions the data into three parts: training data, cross-validation data, and test data. The way these partitions are created varies with the implementation. In some example embodiments, certain training data has been predesignated to be used for training data, cross-validation data, and test data and these designations may be used to partition the data. In other example embodiments, sample job postings are assigned as training data until a preset limit is met, at which point further job postings are assigned as cross-validation data. Then later, if an administrator wishes to test the reliability of the raw industry classification model 400, the sample job postings utilized are designated as test data.); and
outputting, by the industry classification computer program, a custom classification for each of the plurality of startup companies (see at least [0051]: At operation 514, linear classification model fitting is performed. The training data with the selected features form the input for the model fitting process. In addition, three other inputs may be specified. These include regularization types (e.g., L1, L2), regularization constants (e.g., 0.01, 0.1, 0.25, 0.5, 1, 10), and termination tolerances (0.001, 0.01). The linear classification model fitting stages train a model for each of the above parameter configurations independently, and these model files are then scored. In this operation, the feature names and labels are converted to indices representing the row and column indices, respectively, in the matrix used for model fitting.).
Regarding claim 2, Bhadouria further teaches the method of claim 1, wherein the standard industry descriptions are received from one or more third parties or commercial databases (see at least Fig.3 block 302).
Regarding claim 4, Bhadouria further teaches or suggests the method of claim 1, wherein the industry classification computer program converts the standard industry descriptions and the custom industry descriptions to vector representations by using a pre-trained large language model to encode the standard industry descriptions and the custom industry descriptions into high dimensional distributed vector representations (see at least [0056] FIG. 7).
Regarding claim 6, Bhadouria further teaches the method of claim 1, further comprising:
retrieving, by a product/service classification computer program, a plurality of product/service descriptions for the custom industry classifications (see at least [0056] FIG. 7 is a flow diagram illustrating a method 700 of classifying industries for candidate job postings using a raw industry classification model 400, in accordance with an example embodiment. At operation 702, the raw industry classification model 400 is fetched. At operation 704, the feature index IDF and similar industry group files are loaded. At operation 706, one or more job postings corresponding to a single company (e.g., containing the same normalized company name/identification) are obtained. At operation 708, TF-IDF vector calculation is performed for each of the terms in one or more job postings corresponding to the single company. At operation 710, the final TF-IDF vector is used to perform logistic regression classification, which uses the learned coefficients to output the top k k=3) predictions with their prediction scores. At operation 712, the job-posting post-processing component 312 may compute two other derived industry groups from the raw predictions. The first is the top k+ similar industries. This is computed by combining a list of the top k industries from the previous operation as well as every industry similar to each of the top k industries (as identified in an industry similarity table). The second is top k dissimilar industry, which comprises picking the top k industries so that no 2 industries in the set are similar. This helps give breadth to the predicted industries to be assigned to each job posting, with the recognition that is better to have industries that perhaps are not actually relevant to the job posting listed in the job posting than to not have an industry that is relevant to the job posting, as in the latter case a search by a member on the industry would yield no results, whereas in the former the member could always filter out results he or she views as unrelated.);
converting, by the product/service classification computer program, the product/service descriptions and the startup company descriptions to vector representations (see at least [0056] FIG. 7);
performing, by the product/service classification computer program, similarity matching between the vector representations of the product/service descriptions for the custom industry classifications and the vector representations of the startup company descriptions(see at least [0056] FIG. 7); and
outputting, by the product/service classification computer program, one of the product/service descriptions for each industry that the startup company participates based on the similarity matching (see at least [0056] FIG. 7).
Regarding claim 7 The method of claim 1, wherein the startup company descriptions are received by scraping public websites for the startup companies (see at least [0021] FIG. 1 also illustrates a third party application 128, executing on a third party server 130, as having programmatic access to the networked system 102 via the programmatic interface provided by the API server 114. For example, the third party application 128 may, utilizing information retrieved from the networked system 102, support one or more features or functions on a website hosted by a third party. The third party website may, for example, provide one or more functions that are supported by the relevant applications 120 of the networked system 102.).
Claims 8-9, 11, 13-14 essentially recite limitations similar to claims 1-2, 4, 6-7 in form of system thus are rejected for the same reasons discussed in claims 1-2, 4, 6-7 above.
Claims 15-17, 19-20 essentially recite limitations similar to claims 1, 2, 4, 6, 7 in form of non-transitory computer program product, thus are rejected for the same reasons discussed in claims 1, 2, 4, 6, 7 above.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 3, 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bhadouria et al (US 20170300862 A1), in view of Arvindam (US 20210065569 A1).
Regarding claim 3, Bhadouria does not specifically show the method of claim 1, wherein the mapping is provided by a subject matter expert.
However it is customary in the art as shown by Arvindam for subject matter expert to provide knowledge maps (see at least [0043]: A knowledge analysis module is configured for assessing an understanding of the user in a subject matter by comparing the knowledge map created by the user with the knowledge map created by teacher or expert or the gold standard map).
it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include such features taught by Arvindam while implementing the method of Bhadouria in order to provide a reliable reference between the custom industry classification and the standard industry classification.
Claim 10 essentially recites limitations similar to claim 3 in form of system thus is rejected for the same reasons discussed in claim 3 above.
Claim(s) 5, 12, 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Bhadouria et al (US 20170300862 A1), in view of Shacham et al (US 20170344902 A1),
Regarding claim 5, Bhadouria does not specifically show the method of claim 1, wherein the industry classification computer program uses cosine similarly to perform the similarity matching.
However it is customary in the art to do so as shown by Shacham (see at least [0050]: α.sub.w and β.sub.w are weights assigned to w. In an example embodiment, these weights may be based on inverse document frequency. The similarity between a given industry and company may then be calculated as the cosine similarity cos (p.sub.i, p.sub.c); see also [0051]: Thus, for example, the industry “computer networking” may have a cosine similarity of 0.566 to the term “backup”, a cosine similarity of 0.546 to the term “WAN,” a cosine similarity of 0.544 to the term “Router”, a cosine similarity of 0.486 to the term “troubleshoot”, a cosine similarity of 0.485 to the term “VPN,” a cosine similarity of 0.473 to the term “Network security” and a cosine similarity of 0.462 to the term “Virus.” Based on this cosine similarity, some or all of these terms may be selected to be candidates for new industries.);
it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to include such features taught by Shacham while implementing the method of Bhafouria in order to benefit from a standardized technique for finding similar matching information.
Claims 12, 18 essentially recite limitations similar to claim 5 in form of system and non-transitory computer program product respectively, thus are rejected for the same reasons discussed in claim 5 above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Mazzoleni et al (US 20200279171 A1) teach the invention relates generally to the field of computing, and more particularly to data mining and natural language processing (NLP). Word embeddings may be used to capture the meaning of a term in many NLP applications. Word embeddings may map the captured term to a numerical vector representation to find terms with similar meanings associated with similar vector representations. An effective assignment of a word embedding representation to each term may require training on a large corpus of relevant documents, however, a corpus of relevant documents may not be available for specific industries. At 212, the performance is evaluated, and an ending clause is analyzed. At each loop, new documents classified as relevant domain specific data (i.e., industry specific) may be evaluated. Multiple evaluation criteria may be used at each loop. One evaluation criterion may include analyzing the accuracy of the classification algorithm. The classification algorithm analysis may include setting aside part of the training dataset and using the set aside section as a validation set. The analysis may be a statistical approach and may include, for example, the holdout method. The holdout method may cross validate the partitioned training dataset. The holdout process may repeatedly partition the original training dataset into a training dataset and a validation dataset (i.e., cross-validation).
Urmetzer, Florian, Andy Neely, and Veronica Martinez. "Business Ecosystems: Towards a Classification Model." Cambridge Service Alliance (2017).
Abstract
This paper contributes to the business ecosystem literature by offering a classification model, allowing the differentiation of intercompany connections. The problem arose for the researchers that the definition of a business ecosystem lacks separation in the types of connection between companies. Business ecosystems are found to differentiate significantly, starting from loosely coupled to highly regulated and organised company relationships. Some may even result in newly founded business ventures. The authors are proposing a classification model for business ecosystems to allow further classifications in studies.
Bertels, Koen, et al. "Qualitative company performance evaluation: Linear discriminant analysis and neural network models." European Journal of Operational Research 115.3 (1999): 608-615.
Abstract
In this paper, we present a classification model to evaluate the performance of companies on the basis of qualitative criteria, such as organizational and managerial variables. The classification model evaluates the eligibility of the company to receive state subsidies for the development of high tech products. We furthermore created a similar model using the backpropagation learning algorithm and compare its classification performance against the linear model. We also focus on the robustness of the two approaches with respect to uncertain information. This research shows that backpropagation neural networks are not superior to LDA-models (Linear Discriminant Analysis), except when they are given highly uncertain information.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to UYEN T LE whose telephone number is (571)272-4021. The examiner can normally be reached M-F 9-5.
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, Ajay M Bhatia can be reached at 5712723906. 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.
/UYEN T LE/Primary Examiner, Art Unit 2156 16 February 2026