近來，越來越多的使用者會在社群網路，表達對於商品的觀感或互相分享訊息。因此，推薦系統透過上述資源來分析出使用者對商品或是某目標物件的喜好度，以達到推薦給使用者最佳的目標物件。
社群網路可以分為兩類，第一類為使用者會針對目標物件標註喜好度，第二類則不會針對目標物件標註喜好度，上述兩類，互相稱為異質性社群網路。然而，因為第二類社群網路的平台限制，使得使用第一類社群網路建成的推薦系統會比第二類社群網路建成的推薦系統在推薦上有更顯著得推薦效果。
本研究希望以所有使用者為目標，建立旅館推薦系統。而大多數的使用者都有使用第二類社群網路，因此本研究以第二類社群網路的使用者為目標，研發了一套使用異質性社群網路做為輔助資訊的旅館推薦系統。首先，我們設計了一個預測使用者喜好度的模型，來建立針對第一類社群網路使用者的喜好程度矩陣，並將該矩陣視為輔助資訊。然後，利用狄雷克主題模型、分群技術做為兩類社群網路的溝通橋梁，透過分析後的結果能使我們的推薦系統得到更多第一類社群網路的輔助資訊。最後，我們利用輔助資訊和朋友關係網絡，來建構出以第二類社群網路的使用者為目標的推薦模型，在第二類社群網路的使用者尋找旅館時協助做決策，幫助使用者可以找到感興趣的旅館。
本研究將以傳統使用協同式過濾的推薦系統方法及以旅館知名度的推薦系統方法作為參考，來跟我們的推薦系統做效能上的實驗，其結果可以證明使用第二類社群網路的推薦系統，在使用異質性社群網路的輔助資訊後，可以大幅提升效能。

Recently, more and more users share the posts about the products or things in the social media. Therefore, recommendation system can build a rating matrix by analyzing social media data. The system recommends the best items to users.
Then, we divide social media into two categories. The first category collects social media which user gives an item a rating, such as Yelp, Amazon, etc. The second category collects social media which user doesn’t give an item a rating, such as Facebook, Twitter, etc. We regard two categories as heterogeneous social media. Because of the restrictions of the second category, the fact is that recommendation system using the first category more precisely recommends items to a user than recommendation system using the second category.
In this study, we want to build a hotel recommendation system for people. Because most people use social media in the second category, we regard users in the second category as our targets. We developed a hotel recommendation system using heterogeneous social media. First, we design a model for building a rating matrix. This model uses social media in the first category. We regard this rating matrix as auxiliary information of heterogeneous social media. Then, we use LDA and K-means Clustering to build the communication bridges between two categories. These communication bridges make our system extract more auxiliary information of heterogeneous social media. Finally, we design a recommendation model using these communication bridges and friend social network. Through this model, our system recommends the best hotels to users.
We compared the traditional recommendation systems using collaborative filtering method and recommendation system using the popularity of the hotels with our system. The outcomes are expected to evaluate that recommendation system using auxiliary information from the heterogeneous social media is dramatically better than other recommendation systems.

[1] P. Bhattacharya, et al., "Inferring user interests in the Twitter social network," in Proceedings of the 8th ACM Conference on Recommender systems, Foster City, Silicon Valley, California, USA, 2014, pp. 357-360: ACM.
[2] M. J. Pazzani and D. Billsus, "Content-Based Recommendation Systems," in The Adaptive Web: Methods and Strategies of Web Personalization, P. Brusilovsky, A. Kobsa, and W. Nejdl, Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 2007, pp. 325-341.
[3] J. He and W. W. Chu, "A Social Network-Based Recommender System (SNRS)," in Data Mining for Social Network Data, N. Memon, J. J. Xu, D. L. Hicks, and H. Chen, Eds. Boston, MA: Springer US, 2010, pp. 47-74.
[4] B. Berjani and T. Strufe, "A recommendation system for spots in location-based online social networks," presented at the Proceedings of the 4th Workshop on Social Network Systems, Salzburg, Austria, 2011.
[5] S. Bostandjiev, J. O'Donovan, T. H, #246, and llerer, "TasteWeights: a visual interactive hybrid recommender system," presented at the Proceedings of the sixth ACM conference on Recommender systems, Dublin, Ireland, 2012.
[6] S. Zeqian, M. Kwan-Liu, and T. Eliassi-Rad, "Visual Analysis of Large Heterogeneous Social Networks by Semantic and Structural Abstraction," IEEE Transactions on Visualization and Computer Graphics, vol. 12, no. 6, pp. 1427-1439, 2006.
[7] Y. Dong et al., "Link Prediction and Recommendation across Heterogeneous Social Networks," in 2012 IEEE 12th International Conference on Data
41
Mining, 2012, pp. 181-190.
[8] J. Hannon, K. McCarthy, and B. Smyth, "Finding Useful Users on Twitter: Twittomender the Followee Recommender," in Advances in Information Retrieval: 33rd European Conference on IR Research, ECIR 2011, Dublin, Ireland, April 18-21, 2011. Proceedings, P. Clough et al., Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 2011, pp. 784-787.
[9] Y. Kim and K. Shim, "TWILITE: A recommendation system for Twitter using a probabilistic model based on latent Dirichlet allocation," Information Systems archive, pp. 59-77, 2014.
[10] F. Abel, et al, "Analyzing User Modeling on Twitter for Personalized News Recommendations," 2011.
[11] F. Abel, Q. Gao, G.-J. Houben, and K. Tao, "Analyzing temporal dynamics in Twitter profiles for personalized recommendations in the social web," presented at the Proceedings of the 3rd International Web Science Conference, Koblenz, Germany, 2011.
[12] D. Yang, D. Zhang, and e. al, "A sentiment-enhanced personalized location recommendation system," presented at the Proceedings of the 24th ACM Conference on Hypertext and Social Media, New York, NY, USA, 2013.
[13] J. Hannon, M. Bennett, and e. al, "Recommending twitter users to follow using content and collaborative filtering approaches," presented at the Proceedings of the fourth ACM conference on Recommender systems, Barcelona, Spain, 2010.
[14] K. Chen, T. Chen, G. Zheng, O. Jin, E. Yao, and Y. Yu, "Collaborative personalized tweet recommendation," presented at the Proceedings of the 35th international ACM SIGIR conference on Research and development in information retrieval, Portland, Oregon, USA, 2012.
42
[15] J. L. Herlocker, J. A. Konstan, L. G. Terveen, and J. T. Riedl, "Evaluating collaborative filtering recommender systems," ACM Trans. Inf. Syst., vol. 22, no. 1, pp. 5-53, 2004.
[16] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl, "Item-based collaborative filtering recommendation algorithms," presented at the Proceedings of the 10th international conference on World Wide Web, Hong Kong, Hong Kong, 2001.
[17] J. Zhang, P. S. Yu, and Z.-H. Zhou, "Meta-path based multi-network collective link prediction," presented at the Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining, New York, New York, USA, 2014.
[18] X. Yu et al., "Personalized entity recommendation: a heterogeneous information network approach," presented at the Proceedings of the 7th ACM international conference on Web search and data mining, New York, New York, USA, 2014.
[19] S. Zhu, S. Aloufi, J. Yang, and A. El Saddik, "On the learning of image social relevance from heterogeneous social network," Neurocomputing, vol. 210, pp. 269-282, 10/19/ 2016.