本文選題：微博　切入點(diǎn)：產(chǎn)品評(píng)論　出處：《中央民族大學(xué)》2016年碩士論文　論文類型：學(xué)位論文

【摘要】：在網(wǎng)絡(luò)環(huán)境下,信息和數(shù)據(jù)量都在飛速增長(zhǎng),在各種信息中,用戶觀點(diǎn)信息具有著非常重要的作用,在社交網(wǎng)絡(luò)上對(duì)用戶觀點(diǎn)的挖掘可以用于調(diào)查輿情,在電子商務(wù)網(wǎng)站上對(duì)用戶觀點(diǎn)的挖掘可以為商家的產(chǎn)品設(shè)計(jì)和推廣提供有價(jià)值的參考依據(jù),對(duì)用戶觀點(diǎn)的挖掘必須以大數(shù)據(jù)為基礎(chǔ),在用戶觀點(diǎn)信息中含有用戶的潛在興趣,也蘊(yùn)含著用戶的情感狀態(tài)。本文對(duì)基于中文微博產(chǎn)品評(píng)論的用戶觀點(diǎn)挖掘方法展開研究,對(duì)發(fā)現(xiàn)潛在消費(fèi)群體有著非常重要的作用,商家可以據(jù)此制定更有針對(duì)性的產(chǎn)品營(yíng)銷策略,從而真正發(fā)揮出大數(shù)據(jù)挖掘的作用。本文主要包括以下的研究?jī)?nèi)容：1、詳細(xì)探討文本的用戶觀點(diǎn)分析技術(shù),主要包括篇章級(jí)的用戶觀點(diǎn)分類、句子級(jí)的用戶觀點(diǎn)分類以及詞匯的用戶觀點(diǎn)分類等內(nèi)容。2、研究各種分類算法的優(yōu)缺點(diǎn),并通過(guò)實(shí)例比較隨機(jī)森林算法和支持向量機(jī)算法在泛化能力、噪聲魯棒性和不平衡分類上的異同。3、構(gòu)建基于句法依存關(guān)系的微博用戶觀點(diǎn)分析模型和基于文本分類的微博用戶觀點(diǎn)分析模型,并進(jìn)行實(shí)驗(yàn)分析。4、進(jìn)行了系統(tǒng)測(cè)試,主要包括測(cè)試環(huán)境的搭建、基于Hadoop平臺(tái)測(cè)試以及測(cè)試結(jié)果分析等內(nèi)容。本文創(chuàng)新性地提出基于句法依存關(guān)系和文本分類相結(jié)合的中文微博用戶觀點(diǎn)分析算法,測(cè)試結(jié)果表明,算法的正確率和召回率可以接近90%,較改進(jìn)前的算法有了較大幅度的提升。實(shí)現(xiàn)的用戶觀點(diǎn)分析系統(tǒng)具有較強(qiáng)的可靠性和較高的靈活性,易于擴(kuò)展,可以實(shí)現(xiàn)海量微博數(shù)據(jù)的快速篩選,可以將其推廣應(yīng)用于各類社交網(wǎng)絡(luò)和電子商務(wù)網(wǎng)站中。
[Abstract]:In the network environment, the information and the amount of data are increasing rapidly. Among all kinds of information, the user's viewpoint information plays a very important role, and the mining of the user's viewpoint on the social network can be used to investigate the public opinion. The mining of the user's viewpoint on the e-commerce website can provide valuable reference basis for the product design and promotion of the merchant. The mining of the user's viewpoint must be based on big data and contain the potential interest of the user in the information of the user's point of view. This paper studies the mining method of user viewpoint based on Chinese Weibo product review, which plays an important role in discovering potential consumer groups. Based on this, merchants can formulate more targeted product marketing strategies, so that they can really play the role of big data. This paper mainly includes the following research contents: 1, to discuss in detail the user viewpoint analysis technology of the text. It mainly includes user viewpoint classification at text level, user viewpoint classification at sentence level and user view classification of vocabulary. The advantages and disadvantages of various classification algorithms are studied. The generalization ability of stochastic forest algorithm and support vector machine algorithm is compared by examples. The similarities and differences of noise robustness and unbalanced classification. The model of Weibo user viewpoint analysis based on syntactic dependency and the user view analysis model based on text classification are constructed, and the experiment analysis .4is carried out, and the system test is carried out. This paper innovatively proposes a Chinese Weibo user viewpoint analysis algorithm based on syntactic dependency and text classification. The accuracy and recall rate of the algorithm can be close to 90, which is much higher than that of the improved algorithm. The realized user view analysis system has strong reliability, high flexibility and easy to be extended. Weibo data can be quickly filtered and applied to social networks and e-commerce websites.
【學(xué)位授予單位】：中央民族大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2016
【分類號(hào)】：TP391.1

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 謝麗星;周明;孫茂松;;基于層次結(jié)構(gòu)的多策略中文微博情感分析和特征抽取[J];中文信息學(xué)報(bào);2012年01期

2 杜偉夫;譚松波;云曉春;程學(xué)旗;;一種新的情感詞匯語(yǔ)義傾向計(jì)算方法[J];計(jì)算機(jī)研究與發(fā)展;2009年10期

相關(guān)會(huì)議論文 前1條

1 段秀婷;何婷婷;宋樂(lè);;基于PMI-IR算法的Blog情感分類研究[A];第五屆全國(guó)青年計(jì)算語(yǔ)言學(xué)研討會(huì)論文集[C];2010年

，

本文編號(hào)：1644925