-
Book Overview & Buying
-
Table Of Contents
Machine Learning for Emotion Analysis in Python
By :
What does all this add up to? You can make an emotion mining algorithm by making a lexicon with words marked up for emotions. Doing so by extracting the information from a corpus where texts, rather than words, have been marked will probably do better on target texts of the same kind than by using a lexicon where individual words have been marked. They are both time-consuming, labor-intensive activities, but you are going to have to do something like this because any machine learning algorithm is going to require training data. There are numerous minor variants that you can try – stemming, changing the bias, varying the threshold, or using a similarity metric for filling in gaps. They all produce improvements under some circumstances, depending on the nature of the corpus and the task, so it is worth trying combinations of techniques, but they do not produce large improvements. Lexicon-based algorithms form a good starting point, and they have the great advantage of...