An Implementation Of A Dictionary-based Text Sentiment Analysis Algorithm

Internet is a channel by which people express their views,access to informa-tion.Weibo comments,forums and the views of the internet information expressed by other people are becoming more and more,the development of Internet provides favorable conditions for the expression of public opinion information.And these public opinion information will reflect opinions people on a matter,help government find their own work deficiencies in time.But some sensitive negative text informa-tion will affect the social harmony and stability,due to the rapid development of the Internet,some criminals will engage in some online illegal acts.Therefore,the relevant departments need access and grasp on time,analysis the public views which are expressed by Internet users.So establish a emotional analysis system is requried,it can determine the Internet opinion on the public of the emotional polarity automatically.Usually it only consider a single emotional words in some studies of emotional judgment based on the word dictionary-based emotion analysis algorithm,in the recognition of emotional words in the text,but not taking into account the words in the whole sentence of the composition and sentence situation.It easy appear error on the analysis of the results.This paper studies an optimized dictionary-based text emotion analysis algorithm,which also makes a detailed discussion of the emotional computing methods of different sen-tences while recognizing the emotional words.Finally,the design experiments are tested.The main contents of this paper are as follows:1.It analyzes the background and current situation of the text emotion anal-ysis at home and abroad,puts forward the concrete realization process of the text emotion analysis algorithm based on the dictionary,and designs the experiment to test the effect of the algorithm in this paper.2.This paper puts forward a dictionary-based text emotion calculation method which is based on the existing classical emotion method.For different parts of speech,the same part of speech but different emotional tendency of words by setting the weights of different emotions,to choose the appropriate formula according to the characteristics of different patterns after identifying the emotional words.This method can be used to determine the length of text in different length texts such as weibo,forums news,and different areas of text emotion.Finally,it is found that the method proposed in this paper can effectively solve the problem of text emotion analysis,according to the experiments and experimental results.