The overarching goal of the project is develop methods and tools for the identification, analysis and visualization of the flow of information and the distribution of knowledge and attitudes in social media network, in particular blogs.
Sub-goals of the project:
1. Collect and make available for the analysis and visualization of large corpora of texts and links from blogs
2. Analyze data from blogs to model and visualize phenomena like information diffusion and polarization in blog networks
3. To better understand how arguments and opinions are spreading in blog networks
One the project's focuses is the development of methods for analyzing the relationships between the structures of the links connecting the blogs and properties of the texts in the blog posts. We work on methods for identifying clusters in the networks and their properties; identify patterns in the linguistic structures of the blog texts, and characterize the distribution of topics over the network of blogs. For this we develop automated methods for the identification of information structures in texts as well as methods for comparing large volumes of texts. Methods for the analysis of properties of networks of blogs and their development are important in the project as well. The project is developing different types of programs, including applications for the visual analysis of text corpora.
The project has collected large corpora of blogs about climate change and in the domain of surveillance. Through the case studies of these collections the project also contributes to the understanding of how such important topics are discussed in social media.
As the volume of electronically available information continues to grow rapidly, blogs are increasingly being used in the public sphere as sources of information. Thus the blogosphere is becoming an archetypal example of how social networks play a crucial role in mediating the flow of information between individuals and organizations. Interestingly, the blogosphere is both a social network and a network of texts. It is an important research challenge to understand the relationship between the two types of network, and the interplay between the social processes realised in blogs, the flow of information and opinion formation.
When someone is searching for information from social media, it is not sufficient for their search tools to provide relevant inform ation: it is also important for tools to provide a balanced view on a topic and to help users understand more about the source of statements, so they can evaluate them properly.
This project will develop methods and tools to detect, analyse and visualiz e the distribution, flow and development of knowledge and opinions across online social networks. In broad terms, our approach synthesizes techniques from computational linguistics, network analysis and information visualization, which will enjoy a synerg istic relationship with social science theories and models of information diffusion and uptake. The approach includes two principal innovations. Firstly, information is to be characterised by 'key statements', rather than keywords, in order to better capt ure the content of discussions and different opinions. Secondly, the techniques to analyse and visualize online discussions will synthesize data about the network of texts and data about social networks. Tools will be developed with regards to two sets of users: (i) general users who want to access information from the blogosphere and engage in discussions; (ii) media monitors and social scientists who want to analyse information flow and related social processes.