社群媒體是目前各大商業應用獲利的重要媒介，透過使用者的行為分析，企業可以設計出更適合他們的產品。因此，我們必須了解使用者的興趣，以及這些興趣對他們個人而言的重要性。然而，這卻是一項非常艱難的任務，由於大多數使用者不會直接表達他們所熱愛的興趣，必須由他們的社群媒體貼文來推測他們實質的興趣。
現有的研究著重於分析使用者興趣，但是並未考慮興趣的強度以及時間的影響。值得注意的是，人們所喜歡的興趣卻有相當高的機率會隨著時間而改變。在本研究中，我們提出了一套分析使用者興趣的模型，藉由使用者在Twitter上長期發文的資訊，運用時間及次數兩個指標來進行權衡，辨識出他們的個人興趣，並依據其重要性排序。

Many business applications aim to take advantage of social media to target users and increase profit. To achieve this aim, it is necessary to understand the interests that drive users and the personal importance that they assign to each interest; the more importance a given interest has to a user, the more intense or relevant it is. While the end results are desirous, profiling users is a difficult task as users in general are not willing to explicitly reveal information about their interests. It is for this reason that interests must be inferred implicitly from their posts as understanding the users' most intense interests will help in the development of personalized recommendations and advertisements. The existing research in the domain focused on extracting user interests but none so far have considered intensity and the impact of time on expressed interests as factors. It is probable that the way in which a users' interest changes over time can be an ideal indicator of how much a given user likes a given topic. In this study, we propose a model to identify the users' interests and to rank them by importance by leveraging the content of tweets as well the time and frequency that a given user post tweets about their interests on social media.

Introduction .. 1
Related Work .. 4
Overview .. 8
Methodology .. 10
4.1 Pre-Processing .. 10
4.2 Extracting Interest-Relevant Keywords .. 11
4.2.1 Term-Frequency Inverse-Document-Frequency .. 11
4.2.2 Determining relevant keywords in a period of time .. 13
4.2.3 Frequency-Ratio Inverse-Term-Count .. 14
4.3 Tweet Interest Classification .. 15
4.4 Interest Intensity .. 15
4.4.1 Trending Keywords .. 17
4.4.2 Index of Dispersion .. 19
Experiments .. 24
5.1 Experimental Setup .. 24
5.2 Interest Classification Experiment .. 25
5.3 Intensity Ranking Experiment .. 27
5.3.1 Sliding-window size Experiment .. 30
5.4 Survey Discussion .. 30
Conclusion .. 33

[1] Shao-Chen Wu, Elvis Saravia, and Yi-Shin Chen. A dynamic influence keyword model for identifying implicit user interests on social networks. 2013.

[2] Elvis Saravia, Carlos Argueta, and Yi-Shin Chen. Emoviz: Mining the world’s interest through emotion analysis. In Advances in Social Networks Analysis and Mining (ASONAM), 2015 IEEE/ACM International Conference on, pages 753–756. IEEE, 2015.

[3] Pavan Kapanipathi, Prateek Jain, Chitra Venkataramani, and Amit Sheth. User interests identification on twitter using a hierarchical knowledge base. In European Semantic Web Conference, pages 99–113. Springer, Cham, 2014.

[4] John Hannon, Mike Bennett, and Barry Smyth. Recommending twitter users to follow using content and collaborative filtering approaches. In Proceedings of the fourth ACM conference on Recommender systems, pages 199–206. ACM, 2010.

[5] Kailong Chen, Tianqi Chen, Guoqing Zheng, Ou Jin, Enpeng Yao, and Yong Yu. Collaborative personalized tweet recommendation. In Proceedings of the 35th international ACM SIGIR conference on Research and development in information retrieval, pages 661–670. ACM, 2012.

[6] Yang Zhang, Yao Wu, and Qing Yang. Community discovery in twitter based on user interests. Journal of Computational Information Systems, 8(3):991–1000, 2012.

[7] Liangjie Hong, Aziz S Doumith, and Brian D Davison. Co-factorization machines: modeling user interests and predicting individual decisions in twitter. In Proceedings of the sixth ACM international conference on Web search and data mining, pages 557–566. ACM, 2013.

[8] Scott Piao and Jon Whittle. A feasibility study on extracting twitter users’ interests using nlp tools for serendipitous connections. In Privacy, Security, Risk and Trust (PASSAT) and 2011 IEEE Third Inernational Conference on Social Computing (SocialCom), 2011 IEEE Third International Conference on, pages 910–915. IEEE, 2011.

[9] Ying Ding and Jing Jiang. Extracting interest tags from twitter user biographies. In Asia Information Retrieval Symposium, pages 268–279. Springer, Cham, 2014.

[10] Matias Nicoletti, Silvia Schiaffino, and Daniela Godoy. Mining interests for user profiling in electronic conversations. Expert Systems with Applications, 40(2):638–645, 2013.

[11] Wei Shen, Jianyong Wang, Ping Luo, and Min Wang. Linking named entities in tweets with knowledge base via user interest modeling. In Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 68–76. ACM, 2013.

[12] Pavan Kapanipathi, Prateek Jain, Chitra Venkataramani, and Amit Sheth. Hierarchical interest graph from tweets. In Proceedings of the 23rd International Conference on World Wide Web, pages 311–312. ACM, 2014.

[13] Siva Kumar Cheekula, Pavan Kapanipathi, Derek Doran, Prateek Jain, and Amit P Sheth. Entity recommendations using hierarchical knowledge bases. 2015.

[14] Elvis Saravia, Carlos Argueta, and Yi-Shin Chen. Unsupervised graph-based pattern extraction for multilingual emotion classification. Social Network Analysis and Mining, 6(1):92, 2016.

[15] Yoad Lewenberg, Yoram Bachrach, and Svitlana Volkova. Using emotions to predict user interest areas in online social networks. In Data Science and Advanced Analytics (DSAA), 2015. 36678 2015. IEEE International Conference on, pages 1–10. IEEE, 2015.

[16] Kun Xing, Bofeng Zhang, Bo Zhou, and Yucong Liu. Behavior based user interests extraction algorithm. In Internet of Things (iThings/CPSCom), 2011 International Conference on and 4th International Conference on Cyber, Physical and Social Computing, pages 448–452. IEEE, 2011.

[17] Yi-Shin Chen, Yi-Cheng Peng, Jheng-He Liang, Elvis Saravia, Fernando Calderon, Chung-Hao Chang, Ya-Ting Chuang, Tzu-Lung Chen, and Elizabeth Kwan. Concept-based event identification from social streams using evolving social graph sequences. Social Network Analysis and Mining, 5(1):1–16, 2015.

[18] Fernando Calderon, Chun-Hao Chang, Carlos Argueta, Elvis Saravia, and Yi-Shin Chen. Analyzing event opinion transition through summarized emotion visualization. In Proceedings of the 2015 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2015, pages 749–752. ACM, 2015.