DPIR TechTalk: 'Uses of Text and Network Analysis in Politics and the Social Sciences'

The DPIR TechTalks series introduce opportunities offered by computational analysis for social science. For this talk our two speakers were: Professor Stephen Pulman, Department of Computer Science, Signals from text: detecting sentiment, emotion, intent, speculation, risk and deception. Dr Scott Hale, The Oxford Internet Institute: Beyond the individual: Applications of social network analysis for the social sciences Text analysis, including sentiment analysis, named entity recognition and other natural language processing techniques are becoming more widely used in the social sciences to determine metrics and identify trends. Such statistical and machine learning approaches offer a new ways to process bodies of text to yield valuable qualitative and quantitative insights. Professor Stephen Pulman presents a survey of the state of the art in recognising these other dimensions of sentiment in text and describe some practical and some potential applications in finance, politics, and elsewhere. Social network analysis involves a fundamental shift from studying individuals to studying the relationships between people. Dr Scott Hale discusses the applications of social network analysis to large-scale data sources to answer social science questions. One particular application will be to bilingualism and the spread of information between speakers of different languages online. Chair: Radoslaw Zubek, Associate Professor of European Politics, Department of Politics and International Relations