System and method for retrieving results and responses with context based exclusion criteria
AssigneeInternational Business Machines Corporation
InventorsPraveena Viswanathan Panineerkandy, Sujatha B. Perepa, Devasena Vridhachalam
Life of this patent
Estimating the term…
Abstract
A method, system, and computer-usable medium are disclosed for processing a search query that includes context based exclusion criteria. A search query is analyzed as to a subject, topic, noun phrase and query intent. Negative exclusion context and emphasis of the search query are determined with a filter word list to refine the query intent. A knowledge graph is created based on the search query. A set of branches of the knowledge graph is determined from which information is retrieved. Negative intent branches are identified based on the exclusion context and pruning is performed on the negative intent branches from the knowledge graph.
Claims20
A computer-implemented method for processing in sentiment detection comprising:
- analyzing through a knowledge management system that includes a knowledge manager, received search query as to a subject, topic, noun phrase and query intent, wherein the query intent includes positive and negative exclusion criteria;
- during pre-processing of the received search query, removing, by the knowledge manager words in a stop word list during indexing, wherein stop words have a negative intent and context;
- creating, by the knowledge manager a knowledge graph based on the received search query;
- determining, by the knowledge manager a set of branches of the knowledge graph from which information is retrieved;
- identifying, by the knowledge manager negative intent branches based on the positive and negative exclusion criteria;
- applying, by the knowledge manager a parsing mechanism that identifies inclusion and exclusion criteria omitted as stop words and prunes the knowledge graph;
- recognizing, by the knowledge manager exclusion criteria based on the stop words and filtering content based on a filter word list;
- and
- determining, by the knowledge manager a negative exclusion context and emphasis of the received search query with the filter word list to refine the query intent.
A system comprising:
- a processor;
- a data bus coupled to the processor;
- and
- a non-transitory computer-usable medium embodying computer program code, the computer-usable medium being coupled to the data bus, the computer program code used for processing in sentiment detection and comprising instructions executable by the processor and configured for:
- analyzing through a knowledge management system that includes a knowledge manager, received search query as to a subject, topic, noun phrase and query intent, wherein the query intent includes positive and negative exclusion criteria;
- during pre-processing of the received search query, removing, by the knowledge manager words in a stop word list during indexing, wherein stop words have a negative intent and context;
- creating, by the knowledge manager a knowledge graph based on the received search query;
- determining, by the knowledge manager a set of branches of the knowledge graph from which information is retrieved;
- identifying, by the knowledge manager negative intent branches based on the positive and negative exclusion criteria;
- applying, by the knowledge manager a parsing mechanism that identifies inclusion and exclusion criteria omitted as stop words and prunes the knowledge graph;
- recognizing, by the knowledge manager exclusion criteria based on the stop words and filtering content based on a filter word list;
- and
- determining, by the knowledge manager a negative exclusion context and emphasis of the received search query with the filter word list to refine the query intent.
A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for:
- analyzing through a knowledge management system that includes a knowledge manager, received search query as to a subject, topic, noun phrase and query intent, wherein the query intent includes positive and negative exclusion criteria;
- during pre-processing of the received search query, removing, by the knowledge manager words in a stop word list during indexing, wherein stop words have a negative intent and context;
- creating, by the knowledge manager a knowledge graph based on the received search query;
- determining, by the knowledge manager a set of branches of the knowledge graph from which information is retrieved;
- identifying, by the knowledge manager negative intent branches based on the positive and negative exclusion criteria;
- applying, by the knowledge manager a parsing mechanism that identifies inclusion and exclusion criteria omitted as stop words and prunes the knowledge graph;
- recognizing, by the knowledge manager exclusion criteria based on the stop words and filtering content based on a filter word list;
- and
- determining, by the knowledge manager a negative exclusion context and emphasis of the received search query with the filter word list to refine the query intent.
Dependencies are inferred from the claim text.