Reports, Web Pages

Concepts, Relations, Events

Query in English

List relatives of Abu Bakr al-Baghdadi.

Name

Role

Source

Saja al-Dulaimi

wife

Asma al-Dulaimi

wife

Duaa Amid-Ibrahim

sister-in-law

Ibrahim Dulaimi

father-in-law

K-Extractorâ„¢: Convert text into actionable knowledge

Customer Scenario
NEED

Gain specific knowledge from a collection of documents. Identify concepts of interest and their connections, or look up specific events. Enable federated search for both structured and unstructured data resources.

CHALLENGE

Relying on analysts reading through the collection of documents is not feasible anymore. Big Data volumes of unstructured text are impossible to convert to the valuable asset without automation. The automated tools based on shallow concept recognition are throwing away valuable knowledge.

Lymba's Solution
SOLUTION

K-Extractor identifies concepts and events of interest as well as a variety of semantic relations between them. Originally developed to meet the needs of the US Government, it provides event-centric knowledge out of the box. Its powerful inference capabilities allow customizations that satisfy various domain-specific knowledge needs. K-Extractor organizes any unstructured content into a deep semantic structure.

INPUT

Text in virtually any format: from scanned documents in PDF format to plain text files, from popular Office formats to Web pages and social media streams.

OUTPUT

Deep semantic representation: a wide variety of concepts and the semantic relations that link them - in a customizable RDF format.

Key Features

Out-of-the-Box Capability:

Powerful out-of-the-box extraction for a variety of concept types, including people, organizations, events, locations, times, quantities, and many more.

Deep Natural Language Understanding:

Accurate identification of more than 26 semantic relation types to precisely pinpoint the meaning of text, including co-referring links for expressions that refer to the same person or thing.

Configurable & Expandable:

Rapid tailoring for new domains or different knowledge needs by easily adding new concept and relation types.

Check out K-Extractor for more details on the product and its features.