We focus on the influence of emotion on the behavior of Internet forum users and the vitality of online debates. We collected a large set of records describing comments expressed in diverse cyber communities such as blogs, fora and the Digg community. The text was then evaluated using classifiers that were able to estimate emotional valence values. We show that affective interactions do exist in Internet communities and they lead to attractive forces. As a result of collective behaviour there are clusters of comments possessing a similar level of emotional valence that are much longer than they would be if they were created by a random process. The presence of longer clusters of coherent emotional expressions therefore increases the possibility of attaching to this cluster a comment with the same emotion. In this sense there is homophily between participants expressing the same emotional valence. Moreover, we show that emotions are important for community life since the level of the initial emotional valence determines the total length of discussion. The majority of comments possess a negative emotional valence and threads starting from a larger number of negative comments last longer so negative emotions can be treated as a kind of discussion fuel. In the course of time the density of negative comments usually decays and for longer discussions the difference between the emotional level in the beginning and at the end is larger. Since the BBC Forum consists of a number of threads, we can also observe the local activity of a user in a specific thread and his/her average local emotion expressed in this thread. Users can take part in many threads, thus their local and global activities and corresponding emotions can be very different. One can ask: how are a user's emotions expressed in a thread connected to their willingness to be active in a given discussion?  We show that an increase in activity in the discussion of a particular thread is connected with more negative emotions from the user in the thread.

• Legal notice:
  

  Copyright (c) Pielaszek Research, all rights reserved.
  The above materials, including auxiliary resources, are subject to Publisher's copyright and the Author(s) intellectual rights. Without limiting Author(s) rights under respective Copyright Transfer Agreement, no part of the above documents may be reproduced without the express written permission of Pielaszek Research, the Publisher. Express permission from the Author(s) is required to use the above materials for academic purposes, such as lectures or scientific presentations.
  In every case, proper references including Author(s) name(s) and URL of this webpage: https://science24.com/paper/23342 must be provided.

1. q - Ising model on a duplex and a partially duplex clique
2. How are World Events Similar to Contagions?
3. Fluctuation-dissipation relations in physiological data collected during emotional stimulation
4. Emotional clusters observed in Internet communities
5. Transition due to preferential cluster growth of collective emotions in online communities
6. Sentiment and activity analysis of Manchester United football forum
7. Social web sentiment strength detection: methods and issues
8. Online discussions modelled by an evolving Ising-like dynamics
9. Sentiment and activity analysis of Manchester United football forum.
10. Emotional clusters observed in Internet communities
11.  Multidimensional analysis of emotions based on physiological signals
12. Entropy-growth-based model of emotionally charged online dialogues
13. Observations and analysis  of the origins of lognormal distributions of  user post lengths in  Internet discussions
14. Analysis and simulation of emotional states in Internet communities
15. Emotional Agents at the Square Lattice
16. Statistics, emotions and opinions in Digg.com website
17. External bias in the model of isolation of communities
18. Nonequilibrium phase transition due to social group isolation
19. Scaling of human behavior in the portal browsing.
20. Scaling of clusters in a one-dimensional system
21. Scaling of human activity at several Web portal
22. Networks of companies and branches in Poland
23. Log-periodic oscillations in degree distributions of hierarchical scale-free networks