Competitive Intelligence Text Mining: Words Speak

Competitive intelligence (CI) has become one of the major subjects for researchers in the recent years. The present research work is aimed to achieve a part of CI by investigating the scientific articles on this field through text mining in three inter-related steps. In the first step, a total of 1143 articles released between 1987 and 2016 are selected by searching the phrase competitive intelligence in the valid databases and search engines; then through reviewing the topic, abstract, and main text of the articles as well as screening the articles in several steps, the authors eventually selected 135 relevant articles in order to perform the text mining process. In the second step, pre-processing of the data is carried out. In the third step, using nonhierarchical cluster analysis (k-means), 5 optimum clusters are obtained based on the Davies–Bouldin index, for each of which a word cloud is drawn; then the association rules of each cluster are extracted and analyzed using the indices of support, confidence, and lift. The results obtained indicate the increased interest in research works on CI in the recent years and tangibility of the strong and weak presence of the developed and developing countries in formation of the scientific products; further, the results show that information, marketing, and strategy are the main elements of CI that along with other prerequisites can lead to CI and, consequently, the economic development, competitive advantage, and sustainability in the market.