Sentiment analysis 2400-ZEWW929

1. Introductory matters (labs 1).
a. What is sentiment analysis?
b. Levels of sentiment analysis.
c. What is the procedure for performing sentiment analysis and drawing conclusions?

2. Collecting textual data for sentiment analysis (labs 2).
a. Review of web scraping and crawling techniques.
b. Most common technical issues.
c. Ethics and possible legal problems.
d. Review of Python libraries: Selenium and Beautiful Soup with example codes.

3. Textual data preprocessing (labs 3).
a. Tokenization.
b. Stemming.
c. Lemmatization.
d. Stopwords.
e. N-grams.
f. TermFrequence (TF).
g. Inverse Document Frequency (IDF).
h. TF-IDF.

4. Lexicon based sentiment analysis (labs 4-5).
a. Dictionary based approach.
b. Corpus based approach.

5. Unsupervised machine learning methods for sentiment analysis (labs 6).

6. Supervised machine learning methods for sentiment analysis (labs 7-10).
a. Feature types, their extraction and selection for sentiment analysis.
b. Predicting sentiment polarisation vs. emotion detection.
c. Performance evaluation of supervised machine learning algorithms.
d. Review of basic machine learning classification algorithms: logistic regression, naive bayes, k-nearest neighbours, support vector machine and maximum entropy, decision trees, random forests, gradient boosting trees.
e. Neural networks algorithms: RNN, CNN, LSTM, Bi-LSTM, transformers.

7. Possible application of sentiment analysis (labs 11-13).
a. Opinion spam detection.
b. Implicit language detection.
c. Aspect extraction.

8. Students’ presentations (labs 14-15).