In times of mass emergency, vast amounts of data are generated via computer-mediated communication (CMC) that are difficult to manually cull and organize into a coherent picture. Yet valuable information is broadcast, and can provide useful insight into time- and safety-critical situations if captured and analyzed properly and rapidly. We describe an approach for automatically identifying messages communicated via Twitter that contribute to situational awareness, and explain why it is beneficial for those seeking information during mass emergencies. We collected Twitter messages from four different crisis events of varying nature and magnitude and built a classifier to automatically detect messages that may contribute to situational awareness, utilizing a combination of hand-annotated and automatically-extracted linguistic features. Our system was able to achieve over 80% accuracy on categorizing tweets that contribute to situational awareness. Additionally, we show that a classifier developed for a specific emergency event performs well on similar events. The results are promising, and have the potential to aid the general public in culling and analyzing information communicated during times of mass emergency.

The spatial structure of large-scale online social networks has been largely unaccessible due to the lack of available and accurate data about people’s location. However, with the recent surging popularity of location-based social services, data about the geographic position of users have been available for the first time, together with their online social connections. In this work we present a comprehensive study of the spatial properties of the social networks arising among users of three main popular online location-based services. We observe robust universal features across them: while all networks exhibit about 40% of links below 100 km, we further discover strong heterogeneity across users, with different characteristic spatial lengths of interaction across both their social ties and social triads. We provide evidence that mechanisms akin to gravity models may influence how these social connections are created over space. Our results constitute the first large-scale study to unravel the socio-spatial properties of online location-based social networks.

We investigate and evaluate methods for the characterization of social relations from textual communication context, using e-mail as an example. Social relations are intrinsically characterized by the Cartesian product of weights on various axes (we employ valuation and intensity as examples). The prediction of these characteristics is performed by application of unsupervised learning algorithms on meta-data, communication statistics, and the results of deep linguistic analysis of the message body. Classification of sentiment polarity is chosen as the means of linguistic analysis. We find that prediction accuracy can be improved by introducing limited amounts of additional information.

We present several novel minimally-supervised models for detecting latent attributes of social media users, with a focus on ethnicity and gender. Previouswork on ethnicity detection has used coarse-grained widely separated classes of ethnicity and assumed the existence of large amounts of training data such as the US census, simplifying the problem. Instead, we examine content generated by users in addition to name morpho-phonemics to detect ethnicity and gender. Further, weaddress this problem in a challenging setting where the ethnicity classes are more fine grained -- ethnicity classes in Nigeria -- and with very limited training data.

Information extraction systems have been recently proposed for organizing and exploring content in large online text corpora as information networks . In such networks, the nodes are named entities (e.g., people, organizations) while the edges correspond to statements indicating relations among such entities. To date, such systems extract rather primitive networks, capturing only those relations which are expressed by direct statements. In many applications, it is useful to also extract more subtle relations which are often expressed as meta statements in the text. These can, for instance provide the context for a statement (e.g., “Google acquired YouTube on October 2006”), or repercussion about a statement (e.g., “The US condemned Russia’s invasion of Georgia”). In this work, we report on a system for extracting relations expressed in both direct statements as well as in meta statements. We propose a method based on Conditional Random Fields that explores syntactic features to extract both kinds of statements seamlessly. We follow the Open Information Extraction paradigm, where a classifier is trained to recognize any type of relation instead of specific ones. Finally, our results show substantial improvements over a state-of-the-art information extraction system, both in terms of accuracy and, especially, recall.

Gold farming and real money trade refer to a set of illicit practices in massively multiplayer online games (MMOGs) whereby players accumulate virtual resources to sell for “real world” money. Prior work has examined trade relationships formed by gold farmers but not the trust relationships which exist between members of these organizations. We adopt a hypergraph approach to model the multi-modal relationships of gold farmers granting other players permission to use and modify objects they own. We argue these permissions reflect underlying trust relationships which can be analyzed using network analysis methods. We compare farmers’ trust networks to the trust networks of both unidentified farmers and typical players. Our results demonstrate that gold farmers’ networks are different from trust networks of normal players whereby farmers trust highly-central non-farmer players but not each other. These findings have implications for augmenting detection methods and re-evaluating theories of clandestine behavior.

The past year has seen a growing public awareness of the privacy risks of social networking through personal information that people voluntarily disclose. A spotlight has accordingly been turned on the disclosure policies of social networking sites and on mechanisms for restricting access to personal information on Facebook and other sites. But this is not sufficient to address privacy concerns in a world where Web-based data mining tools can let anyone infer information about others by combining data from multiple sources. To illustrate this, we are building a demonstration data miner, GlobalInferencer, that makes inferences about an individual?s lifestyle and other behavior. GlobalInferencer uses linked data technology to perform unified searches across Facebook, Flickr, and public data sites. It demonstrates that controlling access to personal information on individual social networking sites is not an adequate framework for protecting privacy, or even for supporting valid inferencing. In addition to access restrictions, there must be mechanisms for maintaining the provenance of information combined from multiple sources, for revealing the context within which information is presented, and for respecting the accountability that determines how information should be used.

The increased popularity of feature-rich mobile devices in recent years has enabled widespread consumption and production of social media content via mobile devices. Because mobile devices and mobile applications change context within which an individual generates and consumes microblog content, we might expect microblogging behavior to differ depending on whether the user is using a mobile device. To our knowledge, little has been established about what, if any, effects such mobile interfaces have on microblogging. In this paper, we investigate this question within the context of Twitter, among the most popular microblogging platforms. This work makes three specific contributions. First, we quantify the ways in which user profiles are effected by the mobile context: (1) the extent to which users tend to be either fully non-mobile or mobile and (2) the relative activity of the mobile Twitter community. Second, we assess the differences in content between mobile and non-mobile tweets (posts to the Twitter platform). Our results show that mobile platforms produce very different patterns of Twitter usage. As part of our analysis, we propose and apply a classification system for tweets. We consider this to be the third contribution of this work. While other classification systems have been proposed, ours is the first to permit the independent encoding of a tweet’s form, content, and intended audience. In this paper we apply this system to show how tweets differ between mobile and non-mobile contexts. However, because of its flexibility and breadth, the schema may be useful to researchers studying Twitter content in other contexts as well.

What will social media sites of tomorrow look like? What behaviors will their interfaces enable? A major challenge for designing new sites that allow a broader range of user actions is the difficulty of extrapolating from experience with current sites without first distinguishing correlations from underlying causal mechanisms. The growing availability of data on user activities provides new opportunities to uncover correlations among user activity, contributed content and the structure of links among users. However, such correlations do not necessarily translate into predictive models. Instead, empirically grounded mechanistic models provide a stronger basis for establishing causal mechanisms and discovering the underlying statistical laws governing social behavior. We describe a statistical physics-based framework for modeling and analyzing social media and illustrate its application to the problems of prediction and inference. We hope these examples will inspire the research community to explore these methods to look for empirically valid causal mechanisms for the observed correlations.

We perform a sentiment analysis of all tweets published on the microblogging platform Twitter in the second half of 2008. We use a psychometric instrument to extract six mood states (tension, depression, anger, vigor, fatigue, confusion) from the aggregated Twitter content and compute a six-dimensional mood vector for each day in the timeline. We compare our results to a record of popular events gathered from media and sources. We find that events in the social, political, cultural and economic sphere do have a significant, immediate and highly specific effect on the various dimensions of public mood. We speculate that large scale analyses of mood can provide a solid platform to model collective emotive trends in terms of their predictive value with regards to existing social as well as economic indicators.