Solutions

Our natural language processing (NLP) and entity extraction software has uses across many domains.

Natural Language Processing

Idyl E3 can be used as a core component of a larger solution for natural language processing tasks. Many NLP applications such as question and answering systems and sentiment analysis applications require extraction of named-entities in order to function. Idyl E3’s named-entity extraction API allows for integration into NLP pipelines.

Law Enforcement

Some data in law enforcement records are structured data and defined by specifications such as LEXS and NIEM. However, there can be unstructured data inside the structured data as in the Narrative element which is a natural language description of the event. Idyl E3 allows for processing the narrative and extracting entities from it. The entities contained in the narrative can further the agency’s understanding of the record and uncover previously unknown connections.

When used in an Apache NiFi pipeline, Idyl E3 and the Entity Query Language (EQL) can be used to provide alerts when entities meeting specified conditions are found. Along with custom trained entity models for places of interest, weapons, and persons, this toolset provides a powerful means of ingestion and notification capabilities for law enforcement agencies.

Healthcare

Healthcare sees a combination of structured and unstructured data. Patient information such as name, age, and other personal attributes are organized as structured data. Unstructured data can be physician’s notes and assessments. Idyl E3’s entity models trained to identify hospitals and doctors can quickly extract useful information from the text. Soon, Idyl E3 will be able to also identify drugs and dosages.

Legal

Unstructured data is more common than structured data in legal and law environments. Legal proceedings and documents are almost exclusively composed of natural language text. This text commonly contains named entities such as persons and places. The documents are complex and can be long requiring significant time to understand their contents. With Idyl E3 and entity models for persons and places, the time required to develop and understanding of the documents can be shortened. Documents can be organized based upon their contents and archived in a useful fashion.

Scientific Applications

Unstructured text is prevalent in scientific industries such as bioinformatics. With Idyl E3’s custom feature generators, entity models can be trained to identify complex sequences of text such as chemical compounds or genetic codes.