Back to results
US 8,898,163 B2Granted
Real-time information mining
AssigneeInternational Business Machines Corporation
InventorsNilanjan Banerjee, Dipanjan Chakraborty, Anupam Joshi, Sumit Mittal, Seema Nagar, Angshu Rai, Koustuv Dasgupta
Life of this patent
Estimating the term…
Abstract
Embodiments of the invention are related to identifying a user's intent dynamically from at least a set of metadata associated with the user, wherein the set of metadata is associated with a user input, and providing to the user a set of labeled instances on determination of a user's intent, the set of labeled instances being directly related to user's intent, where the set of labeled instances are obtained in real-time from a set of information repositories.
Claims19
A method comprising:
- identifying a user associated with a text expression containing one or more content-indicative words;
- classifying data of the text expression using one or more classifiers, said classifying comprising using the one or more content-indicative words to identify an intent of a user, the intent comprising a semantic meaning of the text expression, the one or more content-indicative words including one or more of:
- at least one temporal keyword, and at least one action keyword;
- and
- on determination of intent of a user based on said identifying and classifying, providing to the user data related to the intent of the user obtained in real-time from a set of information repositories;
- wherein the intent of the user comprises at least one of structured data and unstructured data;
- and
- said classifying comprises employing a Δ-TFIDF based analyzer and at least one classifier selected from the group consisting of:
- a co-occurrence driven analyzer;
- a part-of-speech analyzer;
- a rule-based analyzer;
- a dependency grammar based analyzer.
9 more elements
A system comprising:
- at least a processor;
- and
- a memory;
- wherein said at least one processor and said memory are configured to:
- identify a user associated with a text expression containing one or more content-indicative words;
- classify data of the text expression using one or more classifiers, said classifying comprising using the one or more content-indicative words to identify an intent of a user, the intent comprising a semantic meaning of the text expression, the one or more content-indicative words including one or more of:
- at least one temporal keyword, and at least one action keyword;
- and
- provide to the user data related to the intent of the user obtained in real-time from a set of information repositories;
- wherein the intent of the user comprises at least one of structured data and unstructured data;
- and
- to classify comprises employing a Δ-TFIDF based analyzer and at least one classifier selected from the group consisting of:
- a co-occurrence driven analyzer;
- a part-of-speech analyzer;
- a rule-based analyzer;
- a dependency grammar based analyzer.
13 more elements
A computer program product, comprising:
- a computer readable storage medium having computer readable program code embodied therewith and executable by at least one processor, the computer readable program code comprising:
- computer readable program code configured to:
- identify a user associated with a text expression containing one or more content-indicative words;
- classify data of the text expression using one or more classifiers, said classifying comprising using the one or more content-indicative words to identify an intent of a user, the intent comprising a semantic meaning of the text expression, the one or more content-indicative words including one or more of:
- at least one temporal keyword, and at least one action keyword;
- and
- provide to the user data related to the intent of the user;
- wherein the intent of the user comprises at least one of structured data and unstructured data;
- and
- to classify comprises employing a Δ-TFIDF based analyzer and at least one classifier selected from the group consisting of:
- a co-occurrence driven analyzer;
- a part-of-speech analyzer;
- a rule-based analyzer;
- a dependency grammar based analyzer.
11 more elements
Dependencies are inferred from the claim text.