This dataset contain Khmer script, mixed Khmer-English, and Romanized Khmer (Khmer word written in Latin script) which focus on profanity, food reviews and product reviews that appeared on Facebook.
Each sentence is labeled into one of these class such positive, negative and neutral.