Building sentiment lexicons for all major languages

Sentiment analysis in a multilingual world remains a challenging problem, because developing language-specific sentiment lexicons is an extremely resourceintensive process. Such lexicons remain a scarce resource for most languages. In this paper, we address this lexicon gap by building high-quality sentiment lexicons for 136 major languages. We integrate a variety of linguistic resources to produce an immense knowledge graph. By appropriately propagating from seed words, we construct sentiment lexicons for each component language of our graph. Our lexicons have a polarity agreement of 95.7% with published lexicons, while achieving an overall coverage of 45.2%. We demonstrate the performance of our lexicons in an extrinsic analysis of 2,000 distinct historical figures' Wikipedia articles on 30 languages. Despite cultural difference and the intended neutrality of Wikipedia articles, our lexicons show an average sentiment correlation of 0.28 across all language pairs.