Tutorial 2 - Lorena Recalde

Social Media Data for E-Government. Digital Citizens and their Degree of Interest in Politics 

Abstract: The amount of citizens’ participation and political involvement within democratic societies may vary from one person to another. Similarly, this diversity of political engagement of users might be observed in online platforms such as Twitter. Hence, considering their political-related online activity, they may be cataloged as highly politically active or poorly politically active citizens. The aim of this tutorial is to provide the ICEDEG audience with an overview of the use of social media data for E-Government and data mining strategies to model users’ interests. Specifically, I will present an approach to measure their degree of interest in politics. In the first part of the tutorial, I will introduce the issues related to the supporting role of Online Social Networks (OSNs) in citizenship evolution. Then, we will see techniques employed to model users’ preferences as well as tweets classification and clustering to do so. Considering Twitter as a widely studied platform in digital citizenship, I will present relevant works in the state of the art. In the second part of the tutorial, we will analyze a case study taking the advantage of a demo. Finally, to conclude, I will present emerging aspects, open issues and challenges in this area. 

Rationale

OSNs have shown to be helpful to build a citizen identity for users. For instance, Twitter has proved to be a useful media platform that facilitates forms of political expression for its users. However, considering the extent of content published every second, the dynamic linkage among users and the wide purpose-oriented nature of Twitter, it is difficult to define the degree of political participation or interest of a digital citizen. Being able to solve this issue is of interest in government and society data management, and more generally in recommender systems research. 

Content

Digital citizenship in OSNs (20 min). Digital citizenship is the ability to use technology to obtain political information. Besides, the frequent use of it elicits the online participation of individuals in society [1]. The impact of OSNs in civic engagement, democratic participation, political party supporters interaction and voting is evident in recent years. People enjoy using the Internet and may benefit from it through the opportunity it offers of letting them participate fully in society.

Slides are available at the following link.

  1. Users’ Preferences Model (40 min). In the light of the massive digital information, people are exposed to, they show interest in diverse topics to a greater or lesser extent. Quantifying a user’s degree of interest in certain content, measuring its correlation with his/her preference for another kind of information, finding patterns between preferences for categories, and clustering similar users to detect changes in society (or model different kinds of citizen groups) are all challenging issues, especially in OSNs where the user interests are not static.
  2. Twitter and E-Government, E-Democracy E-Society (15 min). The number of scientific works related to opinion mining and sentiment analysis in Twitter to potentiate government intelligence (or predict elections) is constantly growing. Surveys about political opinion mining and political orientation classification in Twitter are presented in [2,3]. Besides, topics like vote prediction and political ideology monitoring [4,5] are currently being studied.
  3. Social entity recommendation for tweeters (15 min). In the context of recommendation, generally, a system has to discover the interests of the target user in order to have an overview of his/her eventual needs. The Social Web has shown to be one of the richest sources for mining people’s interests, personality, and social interactions. Research works like [6,7] propose methods to solve the feed filtering and ranking problem (tweet recommendation). “Users to follow” recommendation is addressed in [8, 9]. Hashtag hijacking is a common practice in the context of politics. The detection of hijackers to recommend political groups “not to follow them” is analyzed in [5].
  4. Case study (60 min). Twitter has been broadly considered in the literature because of its effect on recent socio-political changes in different countries and regions. By identifying the level of interest in politics of a user we can provide them with meaningful recommendations such as political actors to follow, tweets talking about politics, and political-oriented lists, among others. I present an approach to evaluate the Degree of Interest in Politics “DoIP” of citizens in Twitter [10] depending on the quantity of political-related tweets the user has published. Once this metric is found, it may be used as input to model the user’s political profile.
  5. Summary (30 min). Emerging topics: Consuming biased content in OSNs, Open Issues, and Research challenges. Time for questions and discussion included. 

References

  1. K. Mossberger, C. J. Tolbert, and R. S. McNeal, Digital Citizenship: The Internet, Society, and Participation. The MIT Press, 2007.

  2. E. Mart ́ınez-Ca ́mara, M. T. Mart ́ın-Valdivia, L. A. Uren ̃a-Lo ́pez, and A. Montejo- Ra ́ez, “Sentiment analysis in Twitter,” Natural Language Engineering, vol. 20, no. 1, p. 1–28, 2014.

  3. M. M. Mostafa, “More than words: Social networks’ text mining for consumer brand sentiments,” Expert Systems with Applications, vol. 40, no. 10, pp. 4241 – 4251, 2013.

  4. E. T. K. Sang and J. Bos, “Predicting the 2011 Dutch Senate election results with Twitter,” in Proceedings of the Workshop on Semantic Analysis in Social Media, (Stroudsburg, USA), pp. 53–60, Association for Computational Linguistics, 2012.

  5. L. Recalde, J. Mendieta, L. Boratto, L. Teran, C. Vaca, and G. Baquerizo, “Who you should not follow: Extracting word embeddings from tweets to identify groups of interest and hijackers in demonstrations,” IEEE Transactions on Emerging Topics in Computing, vol. PP, no. 99, pp. 1–1, 2017.

  6. K. Chen, T. Chen, G. Zheng, O. Jin, E. Yao, and Y. Yu, “Collaborative personalized tweet recommendation,” in Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR ’12, (New York, NY, USA), pp. 661–670, ACM, 2012.

  7. S. Berkovsky and J. Freyne, Personalised Network Activity Feeds: Finding Needles in the Haystacks, pp. 21–34. Switzerland: Springer, 2015.

  8. J. Hannon, M. Bennett, and B. Smyth, “Recommending Twitter users to follow us- ing Content and Collaborative Filtering approaches,” in Proceedings of the Fourth ACM Conference on Recommender Systems, RecSys ’10, (New York, NY, USA), pp. 199–206, ACM, 2010.

  9. Y. Liu, X. Chen, S. Li, and L. Wang, “A user adaptive model for followee recommendation on Twitter,” in Natural Language Understanding and Intelligent Applications (C.-Y. Lin, N. Xue, D. Zhao, X. Huang, and Y. Feng, eds.), (Cham), pp. 425–436, Springer International Publishing, 2016.

  10. L. Recalde and A. Kaskina, “Who is suitable to be followed back when you are a twitter interested in politics?,” in Proceedings of the 18th Annual International Conference on Digital Government Research, dg.o ’17, (New York, NY, USA), pp. 94–99, ACM, 2017. 

 

Lorena Recalde is a Ph.D. candidate in the Information and Communication Technologies Web Research Group at Pompeu Fabra University, Barcelona, Spain. She got an M.Sc. in Computer Science with an Advanced Information Processing specialization in 2014, from the University of Fribourg (joint master program with the Universities of Bern and Neuchâtel), Switzerland. Her undergraduate studies were in Computer Engineering (2003-2009) at the Escuela Politécnica Nacional, Quito, Ecuador.Lorena Recalde is a 4th year Ph.D. candidate in the Web Science and Social Computing research group at the Pompeu Fabra University (UPF), Spain. Her main research topics are Recommender Systems, with a special focus on user interest modeling and group detection, Web Mining, Machine Learning and Natural Language Processing. As well as authoring conference and journal publications related to Politics in the Social Web, Lorena has participated as an instructor in i) Group Recommender Systems Seminar, University of Fribourg, Switzerland, 2016; and ii) Women in Tech, Scratch for Girls, UPF, Barcelona, 2016. 

 

 

 

 

 

We're Starting In

Next Event

ICEDEG 2019
24 - 26 April 2019
Quito, Ecuador

About Us

We are ICEDEG - a series of conferences organized annually in the area of eDemocracy & eGovernment.

 

Stay Connected on:

Contacts

For general information about the conference/tutorials/sponsorship, including registration, please contact us at:

  • info (@) edem-egov.org
  • +41 26 300 83 55
  • +41 26 300 97 26
  • Boulevard de Pérolles 90,1700 Fribourg, Switzerland