【目的/意義】情感分析目的是對用戶評論等網絡文本信息進行抽取和情感判斷,為企業、政府等機構提供決策支持。本文旨在分析情感分析提出至今相關研究的主題演化規律,及圖書情報學科進行情感分析的研究特點。【方法/過程】本文基于文獻計量學和可視化的方法,系統梳理國內外情感分析研究發文量、學科、主題及研究熱點規律,并利用Web of science學科類別分析情感分析在圖書情報領域的特點。【結果/結論】本文總結出情感分析研究不僅與社交媒體聯系密切,更融合了計算機科學、語言學、文獻學等多學科知識。圖書情報領域開展情感分析研究有廣泛的應用空間和實際需求,目前正處于快速發展時期,研究者應把握這一研究熱點。

【Purpose/significance】The purpose of sentiment analysis is to extract and evaluate on the user’s comments or other online text information, and to gain the user’s sentiment inclination to provide decision support for enterprises,governments and other institutions. This paper describes the subject evolution law of sentiment analysis, and its research characteristics in information and library science. 【Method/process】The paper systematically summarizes the study of sentiment analysis in domestic and overseas based on the method of bibliometrics and visualization.【Result/conclusion】Combined with the literature reviews and visualization map, this paper summarizes that the sentiment analysis research is not only closely connected with social media, but also combined computer science, linguistics, philology and other multidisciplinary knowledge. As for information and library science, there is a wide range of application and practical needs of sentiment analysis.