從顧客評論探勘知覺定位圖

Mining Perceptual Maps from Consumer Reviews

陳昭宏(Chao-Hung Chen)

指導教授 ： 李瑞庭

國立臺灣大學/管理學院/資訊管理學研究所/碩士(2013年)

https://doi.org/10.6342/NTU.2013.10104

摘要

顧客評論是公司珍貴的資源，因為顧客常分享他們的使用經驗並對各產品特徵提供有用的意見。因此，在這篇論文中，我們提出一個MPM(Mining Perceptual Map)的方法，自動地從顧客評論中建構知覺定位圖與雷達圖，知覺定位圖與雷達圖都是廣泛應用於行銷與商業分析中的視覺化工具。從大量的顧客評論中，建構知覺定位圖與雷達圖，可以降低個人主觀的偏見。本實驗結果可提供智慧型手機業者一些實用的觀察與建議，我們提出的方法可幫助公司定位其產品並制定有效的競爭策略。

關鍵字

情感分析 ； 意見探勘 ； 隱含狄利克雷分佈 ； 知覺定位圖 ； 雷達圖

並列摘要

Consumer reviews are valuable resources for companies since consumers usually share their using experiences on the product or provide useful opinions from various aspects such as different product features. Therefore, in this thesis, we propose a method called MPM (Mining Perceptual Map) to automatically build perceptual maps and radar charts from consumer reviews. Perceptual maps and radar charts are business tools widely used in marketing and business analysis. The proposed method may reduce subjective personal bias since perceptual maps and radar charts are mined from a large number of consumer reviews. The experimental results may provide some practical insight for smartphone companies. Our method can help firms to position new products, and formulate effective marketing and competitive strategies.

並列關鍵字

sentiment analysis ； opinion mining ； Latent Dirichlet Allocation ； perceptual maps ； radar charts

參考文獻

[1] C. Albornoz, J. Plaza and L. Gerv’as, A hybrid approach to emotional sentence polarity and intensity classification, Proceedings of 14th Conference on Natural Language Learning, pp. 153-161, 2010.

[2] A. Bhatnagart and S. Ghose, A latent class segmentation analysis of e-shoppers, Journal of Business Research, Volume 57, Issue 7, pp. 758-767, 2004.

[3] S. Bose and N. Gupta, Customer perception of services based on the SERVQUAL dimensions: A study of Indian commercial banks, Journal of Services Marketing Quarterly, Volume 34, Issue 1, pp. 49-66, 2013.

[4] D. Buscaldi, J. Le Roux, J. J. G. Flores and A. Popescu, LIPN-CORE: Semantic text similarity using n-grams, WordNet, syntactic analysis, ESA and information retrieval based features, Proceedings of Second Joint Conference on Lexical and Computational Semantics, pp. 162-168, 2013.

[5] J. C. Albornoz, L. Plaza, P. Gerv’as and A. D’iaz, A joint model of feature mining and sentiment analysis for product review rating, Proceedings of European Conference on Information Retrieval, pp. 55-66, 2011.