About Named Entity Recognition

OpenText Knowledge Discovery Named Entity Recognition identifies and extracts entities from text. An entity is a word, phrase, or block of information, such as a person's name, an address, a date, or a telephone number.

Named Entity Recognition includes a comprehensive set of predefined entities, for many languages and geographical locations, so that you can extract names, credit card numbers, addresses, and so on. You can also extend Named Entity Recognition by defining your own entities.

You can use Named Entity Recognition to:

  • extract entities from documents and add them to metadata fields before you add the documents to your Content componentindex.

    For example, you might extract company names from your document content and tag the documents with these names. Your front-end application can then use these tags to present a list of companies to your users as search filters.

  • identify personally identifiable information (PII) in your data, so that you can manage this data and conform to regulation such as the General Data Protection Regulation (GDPR).

    TIP: OpenText provides an additional PII grammar package for this purpose. For more information, refer to the Named Entity Recognition Grammars User Guide.

  • perform sentiment analysis. Sentiment analysis identifies positive and negative sentiment in text. For example, you can extract positive and negative comments from product reviews. See Sentiment Analysis.

  • redact sensitive information in text or Knowledge Discovery documents, so that you can conform to data protection standards and use your records for multiple purposes.

Knowledge Discoveryallows you to use Named Entity Recognition in several ways:

  • CFS and NiFi Ingest. You can use Named Entity Recognition to enrich documents during the ingestion process, before you add them to the Content component index. For example, you can extract entities and tag the documents so that is easier to find documents related to a specific person, place, or subject.

    To run Named Entity Recognition as part of the ingestion process, use either Connector Framework Server (CFS) or NiFi Ingest. For more information, refer to the Connector Framework Server Administration Guide, or the NiFi Ingest Getting Started Guide.

  • Named Entity Recognition Server. You can use the Named Entity Recognition Server to extract entities, redact information, and perform sentiment analysis on plain text. See Use Named Entity Recognition Server.

  • Build a custom application using the Named Entity Recognition SDK. OpenText provides Named Entity Recognition SDKs for C, .NET, and Java, so that you can include Named Entity Recognition in your own applications. See Deploy Named Entity Recognition SDK.

For more information about which method to use, see Decide Which Named Entity Recognition Product to Use.