Do more with Tisane Entity Extraction

Detect mentions of people, organizations, locations, filenames, phone numbers, crypto addresses, and more.

Entities are elements of relevance or interest in the text. Tisane extracts both standard entities and those relevant to trust & safety/law enforcement applications.

Standard entities are names of people, their social roles, organizations, places, and so on. We also extract cryptocurrency addresses, bank accounts, credit card numbers, phone numbers, software package names, and more.

Every entity entry is an object made of:

• type - the type of the entity
• name - a standard name, if exists; otherwise, the string that was logged
• subtypes - more detailed additional types
• subtype - the first subtype (for backward compatibility purposes)
• mentions - an array of all detected mentions, with:
  • offset
  • length
  • sentence_index
  • text
• wikidata - a Wikidata ID, if exists