Stopwords and Relevance

The last topic to cover before moving on from stopwords is that of relevance. Leaving stopwords in your index could make the relevance calculation less accurate, especially if your documents are very long.

As we have already discussed in Term-frequency saturation, the reason for this is that term-frequency/inverse document frequency doesn’t impose an upper limit on the impact of term frequency. Very common words may have a low weight because of inverse document frequency but, in long documents, the sheer number of occurrences of stopwords in a single document may lead to their weight being artificially boosted.

You may want to consider using the Okapi BM25 similarity on long fields that include stopwords instead of the default Lucene similarity.