Extraction Output
Semantex™ uses a hybrid extraction model that combines statistical, lexical, and grammatical methods in a single processing pipeline that takes advantage of the strengths of each approach, while minimizing weaknesses inherent in any single technique. This hybrid model enables Semantex to perform exceptionally well as a general-purpose extraction solution while at the same time be flexible to be quickly adapted to specific needs.
ENTITIES
Named Entities are people, organizations, locations and other types with proper names like George Bush, Janya and Buffalo. Janya’s Semantex engine consolidates mentions and attributes of these entities across a document, including pronouns and nominal entities. Nominal Entities are entities unnamed in the text but with vital descriptions or known information that may be associated only through these generic terms such as "the company".
RELATIONSHIPS
Relationships are links between two entities or an entity and one of its attributes. Janya has defined a core set of relations of interest to most users, including personal(such as spouse or parent), contact information(such as address or phone) and organizational (such as employee or founder). Relationships may also be customized to a particular domain or user specification through the Semantex Workbench or by Janya’s Professional Services.
EVENTS
Semantex includes a set of pre-defined events over multiple domains including terrorism and finance. In addition, Semantex considers all semantically rich verb forms as events and outputs the corresponding Subject-Verb-Object-Complement (SVOC) structure accordingly. These general events especially and events overall are more valuable when combined with time and location normalization.
PROFILES
Entity Profiles go beyond simple entity detection to create a single repository of all extracted information about an entity contained within a single document. Entity mentions may be names, nominals (the tall man), or pronouns. Profiles contain any descriptions and attributes of an entity from the text including age, position, contact info and related entities and events.
For more information, see the Semantex product page.
