Thai Language Word Sentiment Classification

We require a Native speaker of Thai, living in Thailand, to label several documents and words for sentiment value (positive, neutral, negative) ex: good->positive, bad->negative , blue->neutral The project has two phases. 1) In a first phase you will classify singular words, for positive, negative or neutral sentiment from a text file. 2) In a second phase, you will classify at the sentence and word level with a web interface. The documents are from social networks (twitter, foruns, blogs, etc) and the work is to be done with a custom web-interface developed by us, with a simple click interface. The documents are in general small in size. The initial task is over a 6000 words document. The size of the second task will depend on the initial task, but should be from 500 to 2000 documents. From current experience on other languages, the current speed of labeling is at 50 to 80 documents per hour when we do not do detailed labeling. Detail labeling might make it 2 to 3 times slower. Total time will depend on the total data for the second phase. All documents and links will be given once the project is accepted.