在社群網路興起的今日，每天都有非常多的訊息及資訊不斷的產生，而且每天都會接收到很多來自於好友以及社團的動態消息、電子郵件及訊息，所以大家通常都會花費許多時間在看大量的訊息及資訊；以臉書(FaceBook)為例，因為臉書(FaceBook)的最新動態包含了許多不一定是使用者想知道的訊息及資訊，如好朋友在他的朋友動態按讚或是好朋友在和使用者無關的社團按讚的動態…等等的資訊不一定全部都是使用者想要得知的訊息。
本文提供了個人化的動態推薦排序以及分類的方式，讓使用者能花費更少的時間看到更多的有用資訊。本篇論文主要提供了個人化計算與朋友的親密程度的計算方式。每天都可經由本系統來蒐集最新的動態以及打卡的資訊，再給予按讚、回覆、分享、打卡及動態內的標記不同的權重值，並參考了Google提出的PageRank和臉書提出的EdgeRank的計算方式且加以修改，再加入客製化的排序方式來進行排序以及分類，使得計算方法更符合使用者的喜好以及有興趣的動態及訊息的呈現。
在好友親密程度的計算上，因人與人之間的關係不太可能短時間之內就有很大的轉變，所以在親密程度上會因為近期之內的較少互動慢慢的遞減，再加上必須避免使用者的朋友隨便的使用互動功能，所以我們參考了TF-IDF的算法來避免這種事情的發生；在排序的部分，會考慮好友的親密程度以及動態的類型來進行排序，動態的類型會因為使用者的不同利用Least Square(最小平方法)的方式來給予使用者的喜好權重分數。

Social network has been very popular in today, And there are a lot of message and information being created every day. We usually spend much time reading a lot of messages and information. Using Facebook as an example, it’s latest statuses include many different messages and information which a user doesn’t want to receive, such as the information that one’s friend gave likes to his friend’s status or gave likes to a group which the user is unrelated to etc, All of this is the information that the user doesn’t want to know.
Our system provides a personalized status recommend action and classification fundionality. It could help users spend less time to get useful information. One of the main functions in our study is to calcute score of friendliness in personalized way. Through the system, we collected the daily latest status and checked into places information. We also gave different weighting values for likes, comments, share, check into places and tag in the status. We referenced and modified the PageRank and the EdgeRank algorithms from Google and Facebook. Moreover, we added personalized sorting method to sort and classify statuses to make the algorithm more suitable for users’ preferences, interested status and messages.
On the friendliness calculation, because the relationship between people is hardly to change in a short time, we make the friendliness to slowly decrease if there’s little interaction between them. To avoid users’ arbitrary interactions to attect the friendliness too much, we referenced the TF-IDF algorithm. For sorting, we will consider both the type of status and the score of friendliness, and apply a least-square approach to determine the weightings for different attesting factors.

1. Popescul, A., et al., Probabilistic Models for Unified Collaborative and Content-Based Recommendation in Sparse-Data Environments, in Proceedings of the 17th Conference in Uncertainty in Artificial Intelligence. 2001, Morgan Kaufmann Publishers Inc. p. 437-444.
2. Hannon, J., M. Bennett, and B. Smyth, Recommending twitter users to follow using content and collaborative filtering approaches, in Proceedings of the fourth ACM conference on Recommender systems. 2010, ACM: Barcelona, Spain. p. 199-206.
3. Ghasemi, T. and A. Abhari, A distributed fuzzy recommendation system, in Proceedings of the 15th Communications and Networking Simulation Symposium. 2012, Society for Computer Simulation International: Orlando, Florida. p. 1-6.
4. Bouillot, F., et al., French presidential elections: what are the most efficient measures for tweets?, in Proceedings of the first edition workshop on Politics, elections and data. 2012, ACM: Maui, Hawaii, USA. p. 23-30.
5. Guy, I., et al., Harvesting with SONAR: the value of aggregating social network information, in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. 2008, ACM: Florence, Italy. p. 1017-1026.
6. Guy, I., et al., Personalized recommendation of social software items based on social relations, in Proceedings of the third ACM conference on Recommender systems. 2009, ACM: New York, New York, USA. p. 53-60.
7. Chen, J., et al., Make new friends, but keep the old: recommending people on social networking sites, in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. 2009, ACM: Boston, MA, USA. p. 201-210.
8. Chin, A., B. Xu, and H. Wang, Who should I add as a "friend"?: a study of friend recommendations using proximity and homophily, in Proceedings of the 4th International Workshop on Modeling Social Media. 2013, ACM: Paris, France. p. 1-7.
9. Levandowsky, M. and D. Winter, Distance between Sets. Nature, 1971. 234(5323): p. 34-35.
10. Zhang, T. and V.S. Iyengar, Recommender systems using linear classifiers. J. Mach. Learn. Res., 2002. 2: p. 313-334.
11. Rendle, S., et al., Fast context-aware recommendations with factorization machines, in Proceedings of the 34th international ACM SIGIR conference on Research and development in Information Retrieval. 2011, ACM: Beijing, China. p. 635-644.
12. Podobnik, V., et al. How to calculate trust between social network users? in Software, Telecommunications and Computer Networks (SoftCOM), 2012 20th International Conference on. 2012.
13. Tchuente, D., et al. A Community Based Algorithm for Deriving Users' Profiles from Egocentrics Networks. in Advances in Social Networks Analysis and Mining (ASONAM), 2012 IEEE/ACM International Conference on. 2012.
14. Jian, J. and C. Qun. A trust-based Top-K recommender system using social tagging network. in Fuzzy Systems and Knowledge Discovery (FSKD), 2012 9th International Conference on. 2012.
15. Yang, Y.H., et al., ContextSeer: context search and recommendation at query time for shared consumer photos, in Proceedings of the 16th ACM international conference on Multimedia. 2008, ACM: Vancouver, British Columbia, Canada. p. 199-208.
16. Marinho, L.B., et al., Improving location recommendations with temporal pattern extraction, in Proceedings of the 18th Brazilian symposium on Multimedia and the web. 2012, ACM: São Paulo/SP, Brazil. p. 293-296.
17. Guy, I., et al., Same places, same things, same people?: mining user similarity on social media, in Proceedings of the 2010 ACM conference on Computer supported cooperative work. 2010, ACM: Savannah, Georgia, USA. p. 41-50.
18. Bucher, T., Want to be on the top? Algorithmic power and the threat of invisibility on Facebook. New Media & Society, 2012.