Topic-based knowledge mining of online student reviews for strategic planning in universities. (February 2019)

• Record Type: Journal Article
• Authors: Srinivas, Sharan; Rajendran, Suchithra
• Is Part Of: Computers & industrial engineering. Volume 128(2019)
• Page Start: 974
• Page End: 984
• Publication Date: 2019-02
• Subjects: Online word-of-mouth -- Text mining -- Sentiment analysis -- Topic modeling -- Voice of the customer -- SWOT analysis; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2018.06.034
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: Uses text analytics of online student review to identify voice of customers. Introduces ensemble topic model to extract topics and label reviews automatically. Integrates four different techniques: topic modeling, opinion mining, RCA, SWOT. Provides efficient & economic performance summary of university and its competitors. Helps university leaders in student-centered strategic planning. Abstract: Over the past few years, studies observe a continuous decline in university enrollment and retention rates resulting in millions of dollars in lost revenue. Past research shows that 95% of the students rely heavily on the positive word-of-mouth to select a college. Especially, with the advent of Web 2.0 tools, students are able to make informed decisions due to increased awareness through online reviews. Academic institutions can also leverage this information to understand and improve their students' perception. The objective of this paper is to identify the current strengths, weaknesses, opportunities and threats (SWOT) of a university by analyzing online student reviews using text analytics. Our proposed approach integrates four different techniques: topic modeling, sentiment analysis, root cause and SWOT analyses. First, we introduce an ensemble of Latent Dirichlet Allocation (E-LDA) topic models to automatically identify the key features (topics) that are predominantly discussed by students and categorize each review sentence into the most related topic. We then detect the opinion associated with each sentence (positive, negative and neutral) using sentiment analysis. Finally, a topic-based opinion summary (TOS) for a university is established to identify its strengths and weaknesses from the students' perspective, and the opportunities and threats are determined by analyzing the TOS of the competitors (or other similar institutions). A case study is used to illustrate the feasibility and application of the proposed approach. The results indicate that the proposed method provides an efficient and economic performance summary of a university and its competitors, and could help its leaders in recruitment and retention efforts.