Extracting keyword-related affective words using the text mining technique

Abstract

Existing affective word extraction methods that are used in affective engineering typically include literature research, questionnaires, and interviews. While these techniques can effectively extract affective words, they are often costly and time consuming. This study proposes an affective word extraction method through text mining, which reduces these costs. The proposed process is as follows. First, web crawling is used to collect text data related to various topics. Then, the natural language processing tool is used to preprocess the collected data. Next, word candidates are obtained through the text mining technique. Lastly, the word candidates are evaluated in terms of their suitability as affective words, and the word selection is finalized. To verify the proposed method, a case study was conducted by considering the keyword “product” and extracting affective words. A comparison with data from a previous study showed that the present method extracts affective words as effectively as existing methods. The time and financial cost that incur through text mining are generally smaller than that involved in literature studies, questionnaires, and interviews. Therefore, the affective word extraction method proposed in this paper exhibits an equivalent performance but reduced time and financial costs compared to existing methods.