CCExtractor Development

Research: How can we identify people, locations, events in a piece of text?


Suppose we want to build a tool that uses subtitles to help the news better. We want to find something important and show case more information or news. Think of it as a Knowledge graph for TV.

The information can easily be found. One Google Search Away or parsing it and going on Wikipedia to scrape it.

The hard part is how can we recognize in the most efficent way something is a Person, Location or Event? We don't want to show useless information that will distract the user. So the Approach has to be fast and accurate! Also the Less network requests the better :) So No API tutorials or anything.


How can we accomplish an accurate and quick identification of Locations, People, Events? Research that and submit it in a PDF or Markdown

Task tags

  • nlp
  • machine learning
  • nltk
  • natural language processing

Students who completed this task

Knob, MajesticHero

Task type

  • assessment Outreach / Research