Sentiment Analysis Technology Research Based On Specific Immune Response

"The sentiment analysis" is also known as "The opinion mining". Emotion is akind of expression of human. The human’s response to the objective things, such asthe joys and sorrows reflects the human perception of objective things. where possible,will be stored in various length of the Chinese text. In the field of informationnetwork, the view expression is a way that people express themselves on the Internet.The subjectivity of social language study is a complicated problem, we’ll get differentunderstanding and interpretation from different angles and levels. The main goal ofthe Sentiment analysis in language is mining useful information from the subjectivesentences. With the development of computer network and the expanding scope ofapplication in people’s life, like twitter, face book and other social networks havebecome a modern important form of the people’s attitudes, opinions and evaluationexpressing. Weibo users are faced with the problem of information overload. In orderto deal with amounts of network resources, a good sentiment analysis system canprecisely process text’s tendency of emotion, and provide more scientific basis forusers. It is a potential personalized service technology.The purpose of this paper is tostudy the emotional polarity information in natural language sentence. In general,emotional words always mean the words which with obvious emotional tendency ofmain body, just like adjective, as well as some adverbs, verb and noun, etc.These information with emotional expressing form a large database, and theamount of data rapidly growing in the database rapidly grows every day. The studyhas a lot of obstacles because of the particularity of Chinese itself and irregularity ofnetwork language. To sum up, these statements have the following features:Network emotional expression is real-time, especially under the circumstance thatmobile terminal technologies development so quickly. Some function in computer yesterday has been replaced by the convenient mobile phone, which makesinformation release more real-time.The irregularity. When net users express the emotion on the Internet, sometimesthe information are not grammatical in order to make their own emotional expressionmore image. But this way does not interfere with communication.They are short andconcise, even like Mars, homophonic personalized language, etc. For example,"886","duang", etc.The ambiguity. Chinese expression is variety, broad and profound. Manylanguages have already become different from their original meaning, especially inthe aspect of network language.The body’s immune system, after one thousand years of evolution, formed acomplete set of defense system against outside invision. The immune system is themean object of this paper using for reference. When bacteria, viruses and otherpathogens intrude into the body, the immune cells in human body will recognize theantigen, and form the antibody that has special structure. The antibodies will make theantigen lose of infection, and destroy it out of the body. Then the formation ofantibodies has memory of this kind of antigen. When the same antigenic determinantof antigen enters the body, the antibodies cells will release a lot of antibody in a shortperiod of time. I want to apply the combination of the antibody and antigenicmechanism to the sentiment analysis, looking forward to achieve a differentclassifying of antigen polarity and intensity based on antigen automatically generatethe corresponding antibody. This model and can also achieve automatic wordsrecogniaing the words and efficient, accurate identification of the text emotion, evenprovide a efficient support for later related work.This article wants to do thefollowing:Introduce a system immune emotion tagging of Chinese essay. In view of theexisting Chinese emotion tagging is not clear, the characteristic of nonspecific. Thisarticle from the perspective of biological immunology, draw the lessons from theimmunological knowledge to design a different method, make the different contextand semantic essay had only one antigenic determinant. They will have specificity determinant. The antigenic determinant in the system generates their own determinantautomatically and reduces the manual participation as far as possible, even improvethe efficiency of resolution, which will lay a good foundation for the next immuneemotional processing.Establish a relatively complete Chinese immune emotional polarity analysissystem. Based on the predecessors’ research learning summary, under the structure ofartificial immune system, I want to introduce the Chinese essay in this sentimentanalysis. When the pretreatment of antigen enter recognition area can be effectiveidentification, and the immune system have stability, robustness, memory andextensibility. Not only can effectively identify the existing antigen, and has the abilityto spontaneously added to the corpus. Finally simulation experiment was carried outon the system, mathematical analysis was carried out on the experimental results. Andthe experimental results show that the artificial immune system on Chinese sentimentanalysis can better emotional polarity analysis, and makes the analysis results to theoptimal.