SeaWord NLP Overview

The purpose of this tutorial is to demonstrate how the SeaWord APIs can be used to extract information from text. In this tutorial we will walk-through:
  • How to use the Summarization API to generate short and long summaries from a meeting transcript

  • How to use the Topic Prediction API to predict topics from a meeting transcript

  • How to use the Action Extraction API to extract and summarize actionable items from a meeting transcript

  • How to use the Cross Lingual NER API to extract entities from a given sentence

  • How to use the Machine Reading API to answer questions from a given text

These APIs are protected by an access token. Please send an email to to get yours!
For these endpoints, please use access_token={token} either in the request webhook or the request header.

Summarization API

The Summarization API uses mBART-50 based summarization models to generate summaries of single turn utterances and full transcripts in multiple languages.

Summarization Tutorial —>

Topic Prediction API

The Topic Prediction API uses a combination of abstractive and extractive techniques to return a list of relevant topics and keywords from a document.

Topic Prediction Tutorial —>

Action Extraction API

The Action Extraction API extracts and summarizes action items from a document.

Action Extraction Tutorial —>

Cross Lingual NER API

The Cross Lingual NER API uses one model to extract 30+ common entities from over 100 languages.

Cross Lingual NER Tutorial —>

Machine Reading API

The Machine Reading API performs extractive machine reading to answer questions given a context text.

Machine Reading Tutorial —>