Natural Language Processing and sentiment analysis with TextBlob: a Python NLP library


When comes to NLP in Python, TextBlob is an ideal module to do text processing and maintain a comparatively moderate accuracy for noun_phrases, tokenization, sentiment analysis, and lemmatization even without a trained and advanced ML model to be avail of.

text version of this tutorial:

Textblob Documentation:

GitHub Repo of this episode:

What is words lemmatization:



Comment List

  • Frank Du
    December 14, 2020

    hi, thank you very much for sharing.can you translate a corpus (not only sentences)?I couldnt success because it doesnt contain only one language.

  • Frank Du
    December 14, 2020

    Thank you so much for uploading this video, it helped me a lot with a research project. Greetings from Mexico.

  • Frank Du
    December 14, 2020

    When you do sentiment analysis you can make a scatter graph (import pyplot as plt) using sentiment and objectivity as the two axises. I also include a red point to show the average (because I'm charting many results for a key word).

    I have found the accuracy of the sentiment analysis a bit weak. Comments that are clearly positive ("great job") come back as a 0 (meaning it didn't understand the text). You can make a new corpus but that's a lot of work. Has anyone else had issues with the accuracy of TextBlob and BeautifulSoup?

Write a comment