Sentiment analysis

Also known as "opinion mining" the sentiment analysis is part of the text analysis.

This automatism tries to figure out which subjective impressions are present in a text (e.g. attitudes, emotions or opinions like). So the sentiment analysis can get an orientation (for/against, like/dislike etc.) of the user to a specific topic ('should wallmart sell guns?')

To improve this, the automatism recieves data from NLP, statistics or some machine learning methods.

Cases

The sentiment analysis is widely applied to VOC. VOC means Voice of the customer and describes the process of capturing the customer's expectations, preferences and aversions.

Examples

Some informations show the subjective impression of a person. For example we have some customer reviews for a juice brand on our webshop without stars or thumbs up/down - just plaintext.

• "I like the taste of strawberries" - like strawberries
• "never again!!! i hate it!!!" - dislike against a/all product/s

References

https://lct-master.org/files/MullenSentimentCourseSlides.pdf

https://en.wikipedia.org/wiki/Sentiment_analysis