What ordinary people mean by places may differ dramatically from what experts consider them to be. This is especially evident in how people talk about places in social media, where 'Los Angeles', for instance, could include areas well outside of the city or even in another county. In order to make best use of the information in social media, we need to understand what people mean when they refer to a place. Social annotations provide valuable evidence for harvesting knowledge about places, e.g., learning their boundaries and relations to other places. However, social annotations are noisy, and this can dramatically distort the learned boundaries. In this paper we propose a method that exploits the distinctive property of social annotations -that it is created by many people -to filter out noise. Using a large data set extracted from Flickr we show that our crowd-based noise filtering method can learn accurate boundaries of places, including vague places.

The structure of a social network contains information useful for predicting its evolution. We show that structural information also helps predict activity. People who are "close" in some sense in a social network are more likely to perform similar actions than more distant people. We use network proximity to capture the degree to which people are "close" to each other. In addition to standard proximity metrics used in the link prediction task, such as neighborhood overlap, we introduce new metrics that model different types of interactions that take place between people. We study this claim empirically using data about URL forwarding activity on the social media sites Digg and Twitter. We show that structural proximity of two users in the follower graph is related to similarity of their activity, i.e., how many URLs they both forward. We also show that given friends' activity, knowing their proximity to the user can help better predict which URLs the user will forward. We compare the performance of different proximity metrics on the activity prediction task and find that metrics that take into account the attention-limited nature of interactions in social media lead to substantially better predictions.

Up-to-date geospatial information can help crisis management community to coordinate its response. In addition to data that is created and curated by experts, there is an abundance of user-generated, user-curated data on Social Web sites such as Flickr, Twitter, and Google Earth. User-generated data and metadata can be used to harvest knowledge, including geospatial knowledge that will help solve real-world problems including information discovery, geospatial information integration and data management. This paper proposes a method for acquiring geospatial knowledge in the form of places and relations between them from the user-generated data and metadata on the Social Web. The key to acquiring geospatial knowledge from social metadata is the ability to accurately represent places. The authors describe a simple, efficient algorithm for finding a non-convex boundary of a region from a sample of points from that region. Used within a procedure that learns part-of relations between places from real-world data extracted from the social photo-sharing site Flickr, the proposed algorithm leads to more precise relations than the earlier method and helps uncover knowledge not contained in expert-curated geospatial knowledge bases.

Large ring cyclodextrins have become increasingly important for drug delivery applications. In this work, we have performed replica-exchange molecular dynamics simulations using both implicit and explicit water solvation models to study the conformational diversity of iota-cyclodextrin containing 14 α-1,4 glycosidic linked d-glucopyranose units (CD14). The new quantifiable calculation methods are proposed to analyze the openness, bending, and twisted conformation of CD14 in terms of circularity, biplanar angle, and one-directional conformation (ODC). CD14 in GB implicit water model (Igb5) was found mostly in an opened conformation with average circularity of 0.39 ± 0.16 and a slight bend with average biplanar angle of 145.5 ± 16.0°. In contrast, CD14 in TIP3P explicit water solvation is significantly twisted with average circularity of 0.16 ± 0.10, while 29.1% are ODCs. In addition, classification of CD14 conformations using a Gaussian mixture model (GMM) shows that 85.0% of all CD14 in implicit water at 300 K correspond to the elliptical conformation, in contrast to 82.3% in twisted form in explicit water. GMM clustering also reveals minority conformations of CD14 such as the 8-shape, boat-form, and twisted conformations. This work provides fundamental insights into CD14 conformation, influence of solvation models, and also proposes new quantifiable analysis techniques for molecular conformation studies in the future.

Knowledge produced online often comes in the form of free-text labels, known as tags, with which users annotate the content they create, such as photos and videos. Increasingly, such content is also georeferenced, i.e., it is associated with geographic coordinates. The implicit relationships between tags and their locations can tell us much about how people conceptualize places and relations between them. However, extracting such knowledge from social annotations presents many challenges, since annotations are often ambiguous, noisy, uncertain and spatially inhomogeneous. We introduce a probabilistic framework for modeling georeferenced annotations and a method for learning model parameters from data. The framework is flexible and general, and can be used in a variety of applications that mine geospatial knowledge from user generated content. Specifically, we study two problems -extracting place semantics and predicting locations of photos from tags -and show that performance of our method is comparable to that of state-ofthe-art approaches. Moreover, we show that combining the two problems leads to a better performance on the location prediction task than baseline.