Hate Speech Detection Using Latent Semantic Analysis (LSA) Method Based On Image

Hate speech are a words, actions which is prohibited because it leads to acts that trigger anarchism and violence attidudes toward other individuals or groups. Ethics in the internet are needed considering that internet is a matter that important use for today's society. However, more side are miss using the internet to spread such kind a hate speech, such as ethnicity, religion and race. The development of a system for detecting hate speech through images is quite rare for now a days. Therefore, this study study is classified to detect whether there is an element of hatred in the image that will be selected. In this final project, the author hopes to make how to classify the element of hate speech in an image performed by the machine learning, which later that machine learning can recognize any kind of hate speech on the image through the existing text. With using Latent Semantic Analysis (LSA) method, we get the result of this research is precision 67%, recall 76.84%, and accuracy 57.9%. After creation of this research, it is hoped the computer can know and classify the existence of hate speech in the image.