本文選題：文本聚類　切入點：K-means　出處：《情報科學》2017年10期 　論文類型：期刊論文

【摘要】：【目的/意義】隨著數(shù)字文本數(shù)量的增加,題名與關(guān)鍵詞作為文本表示方式之一,對文本的揭示作用日益突顯。【方法/過程】基于我國農(nóng)產(chǎn)品品牌評價領(lǐng)域的相關(guān)文獻,對文章的題名與關(guān)鍵詞分別進行了聚類分析,并對兩者的聚類結(jié)果進行了對比,發(fā)現(xiàn)基于題名與基于關(guān)鍵詞的聚類結(jié)果不完全一致,但存在交集,進而將題名與關(guān)鍵詞疊加后進行聚類,經(jīng)過分析認為聚類結(jié)果得到了改善�！窘Y(jié)果/結(jié)論】發(fā)現(xiàn)題名與關(guān)鍵詞對文獻內(nèi)容的揭示程度不同,且均不能良好地概括農(nóng)產(chǎn)品品牌評價領(lǐng)域的文獻內(nèi)容,二者融合后結(jié)果得到改善;同時對今后研究中論文題名與關(guān)鍵詞的選取提出了一些建議。
[Abstract]:[purpose / significance] with the increase of the number of digital texts, title and key words are used as one of the ways to express the text, and the role of revealing the text is increasingly prominent. [method / process] based on the relevant literature in the field of brand evaluation of agricultural products in China, In this paper, the title and key words of the paper are analyzed, and the results of clustering are compared. It is found that the clustering results based on title and keyword are not completely consistent, but there is an intersection between them. Then the title and key words are clustered together, and the clustering results are improved. [result / conclusion] it is found that the title and key words reveal different degrees of literature content. Neither of them can generalize the literature content in the field of agricultural product brand evaluation, and the results of the integration of the two methods are improved. At the same time, some suggestions on the selection of title and key words in the future research are put forward.
【作者單位】： 南京大學信息管理學院;南京大學江蘇省數(shù)據(jù)工程與知識服務(wù)重點實驗室;
【基金】：國家社會科學基金項目(12BTQ037) 國家自然科學基金青年項目(71503121)
【分類號】：G254
，

本文編號：1608074