Open neowangkkk opened 6 years ago
This is a latest emotion detection paper in AAAI 2018 conference (Feb 4-10 upcoming). Just for filing and reference. You don't need to dive into it but can have a quick scan to the abstract if have time.
A very good one from Cornel (I think).
positive/negative/neutral: https://medium.com/@sharonwoo/sentiment-analysis-with-nltk-422e0f794b8 https://www.kaggle.com/ngyptr/python-nltk-sentiment-analysis https://www.learndatasci.com/tutorials/sentiment-analysis-reddit-headlines-pythons-nltk/ https://stackoverflow.com/questions/32879532/stanford-nlp-for-python/40496870#40496870
The two papers of emotion detection are very complicated. And it may not be a practical instruction to do the job. Dr. Yang told me he used dictionary-based method to decode the emotion in text. The dictionary he used is: http://www.saifmohammad.com/WebPages/NRC-Emotion-Lexicon.htm
It would be a bonus if you can explore the emotion classification. I suggest if you can firstly look into ways to do only classifying "positive" and "negative emotion". Any questions, let me know.