Require: Reddit comments.

Ensure: Sentiment label (positive, negative, or neutral).

1: For (each comments in the dataset) do.

2: For (each word in the comments) do.

3: Repersented it with its part of speech using POS tagging.

4: Look up the polarity score of it in a sentiment lexicon dictionary.

5: End for.

6: Compute the average polarity of all the words in the comment.

7: If (average polarity > 0.05) then label the comment with a positive sentiment.

8: Else if (average polarity Î [−0.05, 0.05]) then label the comment with a neutral sentiment.

9: Else label the comment with a negative sentiment.

10: End if.

11: End for.