Research on micro-blog sentiment orientation analysis based on improved dependency parsing

Micro-blog sentiment orientation analysis´s important signification is analyzed firstly, and then proposes a new Improved Dependency Parsing to analyze micro-blog sentiment orientation on the base of analyzing the peculiarity of micro-blog texts. This improved algorithm includes the sentiment orientation analysis of emoticons and punctuation, and analyze sentiment orientation base on the distance to the kernel emotional words in syntax structure. Experimentations show the algorithm has the good applicability and robustness.