Extracting Actions with Improved Part of Speech Tagging for Social Networking Texts

With the growing interests in social networking, the interaction of social actors evolved to a source of knowledge in which it become possible to perform context aware reasoning. The information extraction from social networking specially Twitter and Facebook is on of the problem in this area. To extract text from social networking, we need several lexical features and large scale word clustering. We attempt to expand existing tokenizer and to develop our own tagger in order to support the incorrect words currently in existence in Facebook and Twitter. Our goal in this work is to benefit of the lexical features developed for Twitter and online conversational text in previous works, to design and to develop an extraction model for constructing a huge knowledge based on actions.