# US 10,353,929 B2 · System and method for computing critical data of an entity using cognitive analysis of emergent data > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,353,929 B2 - Document ID: US10353929B2 - Status: Granted - Kind: B2 - Application number: 15/416,476 - Filed: 2017-01-26 - Published: 2018-03-29 - Granted: 2019-07-16 - Earliest claimed: 2016-09-28 - Assignees: Mphasis Limited - Applicants: Mphasis Limited - Inventors: Divya Choudhary; Archisman Majumdar; Vibhav Kamath; Udayaadithya Avadhanam; Jai Ganesh - Web page: https://patentgrep.com/patent/US10353929B2 ## CPC - G06F 16/285: {Clustering or classification} - G06F 16/27: Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor - G06F 40/169: Annotation, e.g. comment data or footnotes - G06F 40/30: Semantic analysis - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract System and method for computing critical data of multiple entities is provided. Elements, from a set of predetermined elements, are assigned to sentences of a first dataset associated with an entity of an industry type. Predetermined elements characterize reputation of entities of various industries. Features determined from sentences of first dataset are categorized into groups. Features of a new dataset of a different entity of said industry type are matched with grouped features and same groups are allotted thereto. Classification rules are applied on matched features. Elements are assigned to sentences of new dataset based on allotted groups, which elements are same as those assigned to first dataset. Reputation scores are generated for said entities by determining positive and negative sentiments from the first and new dataset. Steps of assigning elements and grouping are repeated for datasets associated with entities of different industry types for creating a taxonomy for them. ## Claims ### Claim 1 A method, performed by a tool interfacing with multiple communications channels for retrieving data from multiple data sources associated with multiple entities, for computing critical data of the multiple entities, the tool executing instructions stored in a memory via a processor, the method comprising: assigning, by the tool, an element, from a set of predetermined number of elements, to each sentence of a first dataset associated with an entity of an industry type, the predetermined number of elements being characteristic of reputation of entities of various industries, wherein the element is assigned to each sentence of the first dataset based on cognitive analysis capability developed by the tool using pre-existing dataset associated with entities of different industry types collected over a period of time; categorizing, by the tool, features determined from the sentences of the first dataset into one or more groups, wherein each group represents related Features, corresponding to each assigned element; matching features of a new dataset of a different entity of the said industry type with the grouped features of the said industry type for allotting groups to the matched Features similar to the groups of the grouped features; applying one or more classification rules to the matched features, to generate metadata representing instructions for assigning an element to each sentence of the new dataset, the classification rules being created based on the grouped features and the assigned elements; assigning, by the tool, the element to each sentence of the new dataset based on the applied classification rules, thereby the element assigned to each sentence of the new dataset of the different entity of the said industry type is similar to the element assigned to each sentence of the first dataset of the said entity of the said industry type; and generating, by the tool, a reputation score for the said entities, wherein the reputation score is generated by determining positive and negative sentiments in the first and the new dataset, wherein the assigning of element to the first dataset and feature grouping steps are repeated for datasets associated with entities of different industry types for creating a taxonomy for each of the different industry types, and wherein the pre-existing data in the tool is updated with content of the first dataset and the new dataset. ### Claim 2 (depends on 1) The method of claim 1, further comprising: identifying, by the tool, a sentiment associated with the said entities of the said industry type, wherein the sentiment is identified by determining sentences from the first and the new dataset as representing a positive sentiment and a negative sentiment by matching words in the sentences with words stored in a dictionary identified as positive or negative; and generating the reputation score by computing a percentage of sentences in the first and the new dataset representing a positive and a negative sentiment. ### Claim 3 (depends on 1) The method of claim 1, wherein the set of predetermined elements are pre-configured in the tool. ### Claim 4 (depends on 1) The method of claim 1, wherein the elements comprise at least one member selected from the group comprising of: service or product quality label, social responsibility label, employee friendliness label, technology advancement label, and financial condition label. ### Claim 5 (depends on 1) The method of claim 1, further comprising storing the grouped features in a repository, thereby creating a taxonomy for the different industry types. ### Claim 6 A system, interfacing with multiple communications channels for retrieving data from multiple data sources associated with multiple entities, for computing critical data of the multiple entities, the system executing instructions stored in a memory via a processor, the system comprising: an inference engine, executed by one or more processors, in communication with the processor and configured to: assign an element, from a set of predetermined elements, to each sentence of a first dataset associated with an entity of an industry type, the predetermined elements being characteristic of reputation of entities of various industries; categorize features determined from the sentences of the first dataset into one or more groups, wherein each group represents related features corresponding to each assigned element; match features of a new dataset of a different entity of the said industry type with the grouped features of the said industry type for allotting groups to the matched features similar to the groups of the grouped features; apply one or more classification rules to the matched features to generate metadata representing instructions for assigning an element to each sentence of the new dataset, the classification rules being created based on the grouped features and the assigned elements; assign the element to each sentence of the new dataset based on the applied classification rules, thereby the element assigned to each sentence of the new dataset of the different entity of the said industry type is same as the element assigned to each sentence of the first dataset of the said entity of the said industry type; and generate a reputation score for the said entities, wherein the reputation score is generated by determining positive and negative sentiments in the first and the new dataset, wherein the assigning of element to the first dataset and the feature grouping steps are repeated for datasets associated with entities of different industry types for creating a taxonomy for each of the different industry types, and wherein the inference engine comprises a labelling unit in communication with the processor and configured to: assign the element to each sentence of a first dataset fetched from a data storage based on cognitive analysis capability developed by the system using pre-existing dataset associated with entities of different industry types collected over a period of time; and update pre-existing data with content of the first dataset and the new dataset. ### Claim 7 (depends on 6) The system of claim 6, wherein the labeling unit is configured to assign the element to each sentence of the new dataset based on the generated metadata. ### Claim 8 (depends on 6) The system of claim 6, wherein the predetermined elements are pre-configured in the labelling unit, the predetermined elements comprising at least one member selected from the group comprising of: service or product quality label, social responsibility label, employee friendliness label, technology advancement label, and financial condition label. ### Claim 9 (depends on 6) The system of claim 6, wherein the inference engine comprises a feature grouping unit in communication with the processor and configured to categorize features of the first dataset into one or more groups, the features being determined from each sentence of the first dataset. ### Claim 10 (depends on 6) The system of claim 6, wherein the inference engine comprises a classification unit in communication with the processor and configured to: create the one or more classification rules for allotting the groups to the matched features; and generate metadata representing instructions for assigning an element to each sentence of the new dataset based on the groups allotted to the matched features. ### Claim 11 (depends on 10) The system of claim 10, wherein the labeling unit is configured to assign the element to each sentence of the new dataset based on the generated metadata. ### Claim 12 (depends on 6) The system of claim 6, wherein the inference engine comprises a sentiment identifier in communication with the processor and configured to: analyze the elements assigned to the first and the new dataset to identify a sentiment associated with the said entities of the said industry type as representing a positive sentiment or a negative sentiment, and generates the reputation score for each of the said entities by computing a percentage of sentences in the first and the new dataset representing a positive and a negative sentiment. ### Claim 13 (depends on 6) The system of claim 6, wherein the system is a tool hosted in a cloud computing architecture and delivered to users as Software as a Service (SAAS) over a communication network. ### Claim 14 (depends on 6) The system of claim 6, wherein the predetermined elements are pre-configured in the labelling unit, the predetermined elements comprising at least one member selected from the group comprising of: service or product quality label, social responsibility label, employee friendliness label, technology advancement label, and financial condition label. ### Claim 15 (depends on 6) The system of claim 6, further comprising a repository for storing the grouped features, thereby maintaining a taxonomy for different industry types. ### Claim 16 (depends on 10) The system of claim 10, wherein the classification unit operates in conjunction with the repository for matching features of the new dataset with the grouped features of the first dataset. ### Claim 17 (depends on 15) The system of claim 15, wherein the classification unit operates in conjunction with the repository for matching features of the new dataset with the grouped features of the first dataset. ### Claim 18 (depends on 12) The system of claim 12, wherein the sentiment identifier identifies sentences in the first and the new dataset as a positive sentiment or a negative sentiment by matching words in the sentences with words stored in a dictionary identified as positive or negative, and generates the reputation score. ### Claim 19 A computer program product comprising: a non-transitory computer-readable medium having computer-readable program code stored thereon, the computer-readable program code comprising instructions that, when executed by a processor, cause the processor to: a. assign an element to each sentence of a first dataset fetched from a data storage, the sentence associated with an entity of an industry type, the element being assigned to each sentence from a set of predetermined elements, the predetermined elements being characteristic of reputation of entities of various industries, wherein the element is assigned to each sentence of the first dataset based on cognitive analysis capability developed by the tool using pre-existing dataset associated with entities of different industry types collected over a period of time; b. categorize features of the first dataset into one or more groups, wherein each group represents related features corresponding to each assigned element; c. match the features of the new dataset of a different entity of the said industry type with the grouped features of the said industry type to allot groups to the matched features similar to the groups of the grouped features; d. apply one or more classification rules on the new dataset to the matched features to generate metadata representing instructions for assigning an element to each sentence of the new dataset, wherein the one or more classification rules are created based on the assigned elements and the grouped features; e. generate metadata representing instructions for assigning an element to each sentence of the new dataset of the different entity of the said industry type based on the grouping of the matched features; f. assign an element to each sentence of the new dataset based on the metadata, wherein the element assigned to each sentence of the new dataset of the different entity of the said industry type is same as the element assigned to each sentence of the first dataset of the said entity of the said industry type; g. analyze the elements assigned to the first dataset and the new dataset to identify a sentiment associated with the said entities of the said industry type, wherein the sentiment is identified by determining sentences from the first and the new dataset as representing a positive sentiment or a negative sentiment; and h. generate a reputation score for the said entities based on the identified sentiments, wherein steps a and b are repeated for datasets associated with entities of different industry types for creating a taxonomy for each of the different industry types, and wherein the pre-existing data in the tool is undated with content of the first dataset and the new dataset.