Identifying Gender Bias in Online Crime News Indonesia Using Word Embedding

In the digital era, news portals have become a primary source of information for millions of individuals. This study investigated the potential influence of word choice and gender representation in news on the public's perception of gender, emphasizing its implications for gender equality and human rights. Research has shown that the language used in news reporting can reflect gender bias, highlighting the significance of analyzing gender representation in crime-related news. A word-embedding model was employed to identify and mitigate bias in word representation and ensure fairness in the data analysis. This study aims to enhance our understanding of gender representation in Indonesian crime-related news and to apply word-embedding techniques to identify biases in word representation. The results indicate a potential bias in word embeddings, emphasizing the importance of addressing and mitigating biases in language models to avoid reinforcing unfair stereotypes.