Integrate Tisane Topic Extraction into AWS S3 Storage

Find out the subjects the users are talking about. Classify content by topic. Intelligently deduce geographic region.

Super granular. Supports IPTC, IAB, and Wikidata IDs.

Also known as: theme identification, subject detection, or key topic recognition. Topic extraction determines the dominant topics in the text.

Tisane provides the topics at a document level.

When a particular word has multiple interpretations, the sense of the word must be determined in the current context. For example, Jupiter is a planet and a Roman deity. Whether it's the planet or the deity, depends on the text.

For example, the sentence Juno is the wife of Jupiter refers to the deity. Tisane determines the relevant topics as Roman mythology, supernatural (gods), relationship, and family (since the spousal connection is mentioned).

There are common taxonomy standards that Tisane can provide::

• native - native Tisane topic names; based on standard English terms for the topic. The default standard.
• iptc_code - codes of the IPTC (International Press Telecommunications Council) Media Topics classification - a standard used in the media.
• iptc_description - English descriptions of the IPTC codes.
• iab_code - codes of the IAB (Interactive Advertising Bureau) content taxonomy.
• iab_description - English descriptions of the IAB codes.
• wikidata - Wikidata codes (usually of the form Qnnnnn, e.g. Q123).