本文選題：情感極性 + 網(wǎng)絡(luò)新聞評論��； 參考：《計算機工程與應(yīng)用》2017年01期

【摘要】：網(wǎng)絡(luò)新聞評論情感分析對于互聯(lián)網(wǎng)時代分析輿情、掌握民調(diào)具有重要意義。目前研究聚焦在評論自身的分析而忽略評論間的結(jié)構(gòu)關(guān)系,因此利用該關(guān)系生成評論關(guān)系樹,并基于評論關(guān)系樹建立情感極性判別規(guī)則。將評論經(jīng)過預(yù)處理后,同時采用基于擴展情感詞典和支持向量機兩種方法來進行情感極性分析,動態(tài)擴展了情感詞典,設(shè)計了情感極性分類器。實驗結(jié)果表明,在利用了評論結(jié)構(gòu)關(guān)系之后,兩種方法的分析準確率均較沒利用該關(guān)系之前有了明顯的提升。
[Abstract]:The emotional analysis of network news commentary is of great significance for analyzing public opinion and mastering opinion polls in the internet age. At present, the research focuses on the analysis of the commentary itself and neglects the structural relationship between the comments. Therefore, the relationship is used to generate the comment relation tree, and based on the comment relation tree, the rules of judging the emotional polarity are established. After the comments are preprocessed, two methods based on extended emotion dictionary and support vector machine are used to analyze the emotion polarity, the emotion dictionary is dynamically expanded, and the emotion polarity classifier is designed. The experimental results show that the analytical accuracy of the two methods is significantly higher than that of the former.
【作者單位】： 武漢大學(xué)計算機學(xué)院;
【基金】：國家自然科學(xué)基金(No.61272109)
【分類號】：TP391.1
，

本文編號：1854999