sentiment.ai-package {sentiment.ai}

sentiment.ai: Simple Sentiment Analysis Using Deep Learning

Description

This package uses Google's Universal Sentence Encoder, simplifies all the difficulties, and turns it into a sentiment analysis package.

Main Benefits:

• Tolerates spelling mitsakes

• Not dependent on exactly matching a fixed dictionary

• Requires less pre-processing

• More powerful than dictionary-based methods

Main Drawbacks:

• Requires a lot of RAM

• Can be slow on larger datasets (unless using GPU)

Effectively, if you have a reasonably powerful computer, you can use sentiment.ai as a more flexible, powerful, and modern approach to sentiment analysis.

Author(s)

Maintainer: Ben Wiseman [e-mail] [conceptor]

Authors:

• Steven Nydick [e-mail] (ORCID)

• Tristan Wisner [e-mail]

Other contributors:

• Fiona Lodge [e-mail] [contributor]

• Yu-Ann Wang [e-mail] [contributor]

• Veronica Ge [e-mail] [artist]

• Korn Ferry Institute [e-mail] [funder]

See Also

Useful links:

• https://benwiseman.github.io/sentiment.ai/

• https://github.com/BenWiseman/sentiment.ai