Detecting Emotions in Social-Media Streams

Recently my research and development work within the EMOTIVE project - looking at fine-grained, cross-cultural emotion detection from Twitter datasets - has received considerable national and international mass-media attention. Despite its' relatively low budget the project was an overall success; achieving the highest currently known performance on test-datasets in the world (in terms of F-Measure) and processing tweets at a speed of 1500-2000 tweets per second (on an avg. dual-core processor). The system that was developed detects a range of 8 "Basic Emotions", anger, disgust, fear, happiness, sadness, surprise, and shame and confusion, rather than a variation on the less informative positive / negative sentiment score.

The ontology employed gives a rich linguistic context to the eight emotions through its ability to analyse both ordinary speech and slang. This ability to monitor how the public mood changes over time is particularly useful when assessing what interventions are most successful in dealing with civil unrest or concern. However, potential analysis of tweets with the developed system can range from marketing to personality profiling through computational models which are based on emotions.

To me the project was particularly interesting mainly as it allowed me to further focus on my interests that I had throughout my PhD and also to further delve into very interesting social-media questions, NLP related issues in sparse texts processing (i.e. 140 characters per message vs. traditional NLP on large documents) and Ontology processing applications.

The work has resulted in several conference papers (and at least one Journal paper is on the way, with more in the pipeline). Currently we are continuing work within another EPSRC & DSTL funded project and are in collaboration with several organisations to further explore applications of our fine-grained emotions detection system.

Sports Informatics and Social Collaboration

After a long time, I felt it was about time for a quick update on my Blog. Things have been rather hectic in the last few months. After my torn-ligaments injury, I'm still recuperating, although I've been working for five months as a Research Associate in the field of Sports Informatics, for the computer science department at Loughborough University. This work involves grass-root research of various areas in computer science applications within sport. I am tasked with organising seminars, and a symposium, in addition to initial academic research. There are numerous potential applications, such as the use of image analysis in sports science, virtual reality uses in sports science, coaching applications within sports, or AI / intelligent systems in sports-data management / systems solutions in sports science.

The table below (Lames 2012 - Departmental Presentation) illustrates the two-way relationships between computer science and sports science subjects. Essentially any work at these subject intersections is known as the field of sports-informatics (see the IACSS association website, which is an umbrella association for this type of work).

My task in this research position is to establish research links with international and nation wide research centres, for the department. Loughborough has a strong tradition in sports science with international research excellence (see SSEHS or STI, for instance), and there is a lot of potential applications, for example in team player image analysis based tracking algorithms, or the use of Machine Learning in detection of team-play patterns within the computer science department. To me, an area of most interest are the application of communications / social-media technologies in sports. Recently some work looked at several sporting events, and analysed the social-media UGC (User Generated Content). What type of things people are talking about in regard to sporting events, how the fan-athlete relationship is changing from traditional media, and whether any revealing information is shared (Pegoraro 2010, Kassing and Sanderson 2010). I am especially curious whether Twitter and other social-media contributions may be revealing in relation to for example sport draft picks, line-ups and team-play / strategy changes (in other words, these are problems of talent detection and coaching). There is also more work to be done in investigating fan-athlete communications, such as predicting how likely an athlete / celebrity is to respond with a direct message to fans, identifying fans, and classifying them based on the dynamics of interactions, or correlating match tracking data with social media contributions, since this type of work has not seen much research.

I am also beginning a new RA position in the All-in-One project (funded by the EPSRC) at Leicester University. This is a project looking at single infrastructure provision, and its technological and scientific feasibility within a 100 years from now. This progress is motivated by climate change, cost reductions, and efficient use of utilities (see this working paper, for a basic introduction). My main task in this project is to work on a collaborative web-based (web 2.0 / social-media type) system that facilitates collaboration and sharing within an academic and also a wider citizen-science community. This is an interesting area of work, with various problems, such as: how to design a system that facilitates efficient social, web based collaboration of many individuals?; or how to attract and maintain an active user-base of contributors and collaborators on the web based system? There is some very interesting research work in this area that was done within the Climate CoLab project of MIT's Collective Intelligence Centre, and for example the CSCW conference, contains highly relevant research contributions that help answer the questions, above. My work, within the All-in-One project involves the deployment of a collaborative system / processes based on the evaluation of prior academic research. Some of my work within my PhD, such as the design of the Newsmental system, is relevant to this, and it will be interesting to put the entire concept of collective intelligence into practice, within a larger scale project, such as this one.

References:

• [Pegoraro, A. Look Who's Talking - Athletes on Twitter: A Case Study International Journal of Sport Communication, 2010, 3, pp. 501-514]
• [Kassing, J. W. & Sanderson, J. Fan-Athlete Interaction and Twitter Tweeting Through the Giro: A Case Study International Journal of Sport Communication, 2010, 3, pp. 113-128]