近年來，隨著網際網路的發展，消費者能夠在消費之後，在網路平臺上面發表自己對於此次消費的滿意程度，並留下評分供有需求的使用者參考。
本研究目的在於觀察旅店的評論內容和顧客傾向中的不一致性，評論文本內容普遍存在兩個典型的特徵，星星和評論文本內容。評論的文本內容提供了文字用以解釋給分說明，當星星和評論內容對稱(即星等和內容一致)時，會在消費者閱讀購物經驗中加深印象，且提升價值；反之，當評論內容的不確定性提高的時候，使得消費者失望和苦惱，對於消費者和企業，線上評論系統的價值也降低了。
本研究以lexicon-based的方法，不用透過人工標注的方式得到評論的極性，檢查評論當中所存在的不一致性。目的是要過濾評論文本內容傾向和使用者評分傾向不一致的評論，以提高評論資料的可信度。

Since the rapid development of internet and technology, nowadays people are used to check online reviews before they purchase products. On the other head, people can write comments or reviews to express their opinions by any device which is able to access the Internet.
The purpose of our research is to observe and analyze the nonaligned travel reviews. There are two specific features in each review including the ratings and the context of reviews. The context of the review contains the description which describes the reasons for the ratings. When the rating of the review is aligned with the content of the review, the review is more persuasive and impressed. However, if the polarity of the content is not aligned with the ratings, the uncertainty grows and it makes readers confused. The results will lead to mutual disadvantage for both customers and producers and leave bad impression on the online review platform.
Our research employs the lexicon-based method to avoid from annotating by manual work and finds out the reviews that are not aligned. The aim is to filter out the nonalignment reviews so as to enhance the credibility of the review system.

Amabile, T. (1983). Brilliant but cruel: Perceptions of negative evaluators. Journal of Experimental Social Psychology, 19(2), pp.146-156.

Deshpande, S., Rathi, J., Gandhi, S., Shinde, M., & Deshmukh, V. (2015). Sentiment Analysis Tool using Cosine and Jaccard Implementation. International Journal of Computer Applications, 115(12), 17-19. doi:10.5120/20204-2460

Geierhos, M., Bäumer, F. S., Schulze, S., & Stuß, V. (2015). Filtering Reviews by Random Individual Error. Current Approaches in Applied Artificial Intelligence, 305-315. doi:10.1007/978-3-319-19066-2_30

Ku, Lun-Wei and Hsin-Hsi Chen (2007). Mining Opinions from the Web: Beyond Relevance Retrieval. Journal of American Society for Information Science and Technology, Special Issue on Mining Web Resources for Enhancing Information Retrieval, 58(12), pages 1838-1850

Liu, N., He, Y. X., He, F. Y., Peng, M., & Liu, J. B. (2013). The Method of Chinese Opinion Sentence Extraction and Polarity Identification Based on Sentimental Elements. Advanced Materials Research, 765-767, 1406-1410. doi:10.4028/www.scientific.net/amr.765-767.1406

MacQueen ,J. B. (1967): "Some Methods for classification and Analysis of Multivariate Observations", Proceedings of 5-th Berkeley Symposium on Mathematical Statistics and Probability, Berkeley, University of California Press, 1:281-297

Mudambi, S.M., Schuff, D., Zhang, Z.: Why aren’t the stars aligned? an analysis of online review content and star ratings. In: Proceedings of the 47th Hawaii International Conference on System Sciences (HICSS), pp. 3139–3147. IEEE (2014)

Pang, Bo, et al. “Thumbs up?” Proceedings of the ACL-02 Conference on Empirical Methods in Natural Language Processing - EMNLP '02, 2002, doi:10.3115/1118693.1118704.

Trindade, L., Hui Wang, Blackburn, W., & Rooney, N. (2013). Effective sentiment classification based on words and word senses. 2013 International Conference on Machine Learning and Cybernetics. doi:10.1109/icmlc.2013.6890481

Utz, S., Kerkhof, P., and Bos ,J. van den, “Consumers Rule: How Consumer Reviews Influence Perceived Trustworthiness of Online Stores,” Electronic Commerce Research & Applications, 11(1), 2012, pp. 49-58

Yu , Liang-Chih, Lee ,Lung-Hao, Hao ,Shuai, Wang ,Jin, He ,Yunchao, Hu ,Jun, Lai, K. Robert, and Zhang ,Xuejie. 2016. Building Chinese affective resources in valence-arousal dimensions. In Proceedings of NAACL/HLT-16, pages 540-545.

K-means clustering. (2017). En.wikipedia.org. Retrieved 11 July 2017, from https://en.wikipedia.org/wiki/K-means_clustering

謝鎮宇, “意見探勘在中文評鑑語料之應用”, 國立交通大學碩士論文, 99.

沈信佑，“劇本文件探勘與廣告推薦之研究”, 國立台灣師範大學資訊工程所碩士論文, 105

邱鴻達，“意見探勘在中文電影評論之應用”, 國立台灣師範大學資訊工程所碩士論文, 100