International Business Machines Corporation

Benjamin J Bachman

William Scott Spangler

Meenakshi Nagarajan

Olivier Lichtarge

Angela D Wilkins

Lawrence A Donehower

Sam J Regenbogen

Curtis R Pickering

Ying Chen

Linda Ha Kato

Ioana Roxana Stanoi

Jacques Labrie

Maria E Terron-Diaz

Anbu Karani Adikesavan

A method, system, and computer program product for event detection using roles and relationships of entities are provided in the illustrative embodiments. A training event and a set of entities participating in the training event are identified in a training data. For a first entity in the set of entities, a first role occupied by the entity in the event is determined. A behavior attribute is assigned to the first role. A relationship of the first role with a second role corresponding to a second entity in the set of entities is determined. An event rule is constructed to detect an event corresponding to the training event in new data and comprising a plurality of roles, behavior attributes, and the relationship. The plurality of roles includes the first role and the second role, and the plurality of behavior attributes includes the behavior attribute assigned to the first role.

Event detection using roles and relationships of entities

A new information in a language and relating to a subject matter domain is parsed into a constituent set of complete grammatical constructs. In a subset of the complete grammatical constructs, a set of linguistic styles of the language is identified according to a subset of a set of word-style associations related to the language and independent of the subject matter domain. A first weight is assigned to a first linguistic style and a second weight to a second linguistic style from the set of linguistic styles. A first intention information is mapped to the first style using a first style-intention rule, and a second intention information to the second style using a second style-intention rule. A complete grammatical construct in the subset is tagged with the first intention information responsive to a weight associated with the first intention information exceeding an intention selection threshold.

Intention detection in domain-specific information

Growing community of inventors

Houston, TX, United States of America

Benjamin J Bachman