Text mining for better insights

Text mining for better insights

January 17, 2019 DATAcated Challenge 0

Text mining is the process of analyzing (usually unstructured) text in order to get useful insights. Typical tasks in text mining include:

  • Data retrieval, cleaning and shaping
  • Pattern recognition
  • Document clustering
  • Sentiment analysis
  • Relationship, fact and event extraction
  • Disambiguation – i.e. whether word “Excel” means the Microsoft product or the good classic cartridge razor of Gillette

Typical examples of text mining for solving business cases are:

  • Identifying most frequent issues from customer complaints. Ideas for new products could be generated too
  • Analyzing open ended questions in marketing surveys
  • OCR recognition, classifying and booking invoices
  • Review of various media and extracting articles that contain key words

It could be also combined with other quantitative data for better results. For instance: predicting what would be the effect over sales of a negative/positive article about our product in certain media.

By: Kolyu Minevski

 

Leave a Reply

Your email address will not be published. Required fields are marked *