NLTK can be used to remove stopwords. Technique to remove common words(and their plural versions) from a string
However, is there a library contain the "common words" such as "near", "science", and thus select the "uncommon words" such as "attractor" etc. ?