Analysis And Research Of Sentiment Text Data Of Mobile Users Based On Crowd Sensing

The development of Internet and the rapid popularization of mobile intelligent devices promote the development of crowd sensing technology.In addition,text mining is also the trend of development.In this paper,the sentiment text data of the mobile user are analyzed and studied in the context of crowd sensing.The main work accomplished includes the following aspects.1.A sensing node selection mechanism based on sentiment text data screening is proposed.By combining with the existing sensing node selection method and the research content of this paper,the previous node selection is improved.The improvement is mainly based on the analysis of the sentiment text data.The data carried by the node is screened before the node is selected,and then the selection of the sensing node is realized with genetic algorithm.The feasibility and effectiveness of the method are verified by experiments.2.Preprocessing of sentiment text data.Combining with the characteristics of data fragmentation in the crowd sensing environment,the main work is done in this part and it consists of two aspects : 1)the process design of the integration of fragmented text;2)the definition of the clustering rules of the integration layer.In the process of integrated process design,the work of each layer is introduced.In the definition part of clustering rules,the semantic analysis of sentences and the summary of clustering rules are made in combination with concrete examples.3.A feature selection of sentiment text based on CHI and Information Gain is proposed.In view of the problems that the existing feature selection methods are mainly aimed at the common text,the existing text feature selection methods are improved by combining with the research content of this paper.Experimental results show that the proposed method is effective in affective text classification.4.Tendency analysis of sentiment text.Based on the acquisition of the previous sentiment text data and the improved feature selection algorithm of sentiment text,the sentiment tendency of the sentiment text is judged according to the selected sentiment feature words.The method is mainly based on the emotional polarity table,and the classification results are explained.