GSLIS Assistant Professor Jana Diesner will deliver a keynote speech at the Daegu Gyeongbuk International Social Network Conference (DISC) 2013, which will be held in South Korea from December 12-14. The theme of the conference is "Knowledge Network Analysis in the Emerging Big Data Research."
Diesner’s keynote, "Words and Networks: Relevance of methodological choices for practical applications," will address the following:
Text mining techniques can be used to construct and enhance social network data. This approach is particularly useful for obtaining network data on hard-to-access groups and organizations, and to enhance given social network data with the information authored or disseminated by network participants, respectively. These techniques require the selection of appropriate methods, algorithms and parameter settings. The impact of these choices on the resulting data and findings can be strong, but is hardly understood.
In the first part of the talk, I present our findings from addressing this problem: We applied four common relation extraction methods—from fairly qualitative to fully automated, machine-learning based ones—to large-scale, open-source corpora from the business, science and geopolitical domain, and compared the retrieved networks. I will address common agreements and disagreements about network structure and behavior depending on the methodological choices, and show how these methods can be combined to gain a more robust and comprehensive understanding of a network.
In the second part of the talk, I put these findings in an application context: I present on our recent work on developing a theoretically-grounded, empirical and computational methodology and technology for studying the impact of social justice documentaries on individuals, groups and society. We approach this task by constructing, analyzing and monitoring socio-semantic networks that represent relevant stakeholders and the information they share and disseminate. I show comparative findings from applying this solution to multiple documentaries and report on the utility of this work to documentary (impact) producers.
Diesner also will chair the panel "Mixed Network Analysis" and present a workshop, "Semantic Network Analysis using ConText." ConText, a data analysis tool created by Diesner and her lab, supports the construction of network data based on text data and meta data, and the joint analysis of text data and network data.