الفهرس | Only 14 pages are availabe for public view |
Abstract It is a non-trivial task to maintain language neutrality, especially in Wikipedia biographical articles about men and women given the open, crowdsourced nature of Wikipedia. For that purpose, an automatic system is developed to detect gender-based bias and its intensity in Wikipedia biographical articles about men and women. Based on the tenents of the Linguistic Category Model (LCM) about the abstract biased language and concrete neutral language, the study uses the English Multi-perspective Question Answering (MPQA) Subjectivity lexicon and its Arabic translation to detect verbs, adjectives, nouns, and adverbs in a corpus of 1600 Wikipedia articles covering both men and women from different professions in both English and Arabic. The current study tries to answer the following questions: 1) What is the evidence of gender bias in each of the Arabic and English male and female articles? 2) Which of the Arabic and English Wikipedia male and female datasets show more intense gender bias? and 3) How does the intensity of gender bias change when comparing specific professions between the Arabic and English male and female datasets? The results show evidence of intense gender bias particularly in the analyzed Arabic male Wikipedia articles compared with Arabic female and English male and female Wikipedia articles. It has also been found that Arabic Wikipedia articles about male politicians are more intensely biased compared to sportsmen and male writers. |