Efficient Calculation of Bigram Frequencies in a Corpus of Short Texts

This addresses a computational bottleneck for researchers or practitioners working with short-text data, but it is incremental as it modifies an existing method for a specific scenario.

The authors tackled the problem of calculating bigram frequencies in corpora of short texts, showing that an existing efficient method is unsuitable and proposing a simple alternative with the same computational complexity that provides exact counts instead of approximations.