Android Based Hate Speech Search Applications Using TF-IDF Algorithm and Vector Space Models

Abstract

The freedom in the use of social media becomes a common means in voicing opinions and expressions by
issuing words or phrases of blasphemy (Hate Speech) in social media like Facebook. Hate speech and blasphemous
words are easily spread across social and general media not found that have transgressed the limits and can trace the
unrest. To find out what the users of social media especially Facebook we have issued words or phrases of blasphemy
on the basis of this report occurred. The study was conducted by using TF-IDF weighting algorithm and VSM (Vector
Space Model) calculation, while the data used was the 30 users post using Graph API. This research resulted in search
with recall value: 2/2 = 1 which means the result of the relevant document and the precision value of the search result
is 2/13 = 0,154 which means result also raises irrelevant search result. Percentage of hate speech equivalence value
was 43%.