We present a series of visual information extraction experiments using the Faces of Wikipedia database - a new resource that we release into the public domain for both recognition and extraction research containing over 50,000 identities and 60,000 disambiguated images of faces. We compare different techniques for automatically extracting the faces corresponding to the subject of a Wikipedia biography within the images appearing on the page. Our top performing approach is based on probabilistic graphical models and uses the text of Wikipedia pages, similarities of faces as well as various other features of the document, meta-data and image files. Our method resolves the problem jointly for all detected faces on a page. While our experiments focus on extracting faces from Wikipedia biographies, our approach is easily adapted to other types of documents and multiple documents. We focus on Wikipedia because the content is a Creative Commons resource and we provide our database to the community including registered faces, hand labeled and automated disambiguations, processed captions, meta data and evaluation protocols. Our best probabilistic extraction pipeline yields an expected average accuracy of 77\% compared to image only and text only baselines which yield 63\% and 66\% respectively.

Graph clustering or community detection constitutes an important task forinvestigating the internal structure of graphs, with a plethora of applications in several domains. Traditional tools for graph clustering, such asspectral methods, typically suffer from high time and space complexity. In thisarticle, we present CoreCluster, an efficient graph clusteringframework based on the concept of graph degeneracy, that can be used along withany known graph clustering algorithm. Our approach capitalizes on processing thegraph in a hierarchical manner provided by its core expansion sequence, anordered partition of the graph into different levels according to the k-coredecomposition. Such a partition provides a way to process the graph inan incremental manner that preserves its clustering structure, whilemaking the execution of the chosen clustering algorithm much faster due to thesmaller size of the graph's partitions onto which the algorithm operates.

Influence maximization problem is to find a set of seed nodes in a social network such that their influence spread is maximized under certain propagation models. A few algorithms have been proposed for solving this problem. However, they have not considered the impact of novelty decay on influence propagation, i.e., repeated exposures will have diminishing influence on users. In this paper, we consider the problem of influence maximization with novelty decay (IMND). We investigate the effect of novelty decay on influence propagation on real-life datasets and formulate the IMND problem. We further analyze the problem properties and propose an influence estimation technique. We demonstrate the performance of our algorithms on four social networks.

Contemporary machine translation systems usually rely on offline data retrieved from the web for individual model training, such as translation models and language models. In contrast to existing methods, we propose a novel approach that treats machine translation as a web search task and utilizes the web on the fly to acquire translation knowledge. This end-to-end approach takes advantage of fresh web search results that are capable of leveraging tremendous web knowledge to obtain phrase-level candidates on demand and then compose sentence-level translations. Experimental results show that our web-based machine translation method demonstrates very promising performance in leveraging fresh translation knowledge and making translation decisions. Furthermore, when combined with offline models, it significantly outperforms a state-of-the-art phrase-based statistical machine translation system.

Entity search is to retrieve a ranked list of named entities of target types to a given query. In this paper, we propose an approach of entity search by formalizing both context matching and category matching. In addition, we propose a result re-ranking strategy that can be easily adapted to achieve a hybrid of two context matching strategies. Experiments on the INEX 2009 entity ranking task show that the proposed approach achieves a significant improvement of the entity search performance (xinfAP from 0.27 to 0.39) over the existing solutions.

Online social networking sites have become popular platforms on which users can link with each other and share information, not only basic rating information but also information such as contexts, social relationships, and item contents. However, as far as we know, no existing works systematically combine diverse types of information to build more accurate recommender systems. In this paper, we propose a novel context-aware hierarchical Bayesian method. First, we propose the use of spectral clustering for user-item subgrouping, so that users and items in similar contexts are grouped. We then propose a novel hierarchical Bayesian model that can make predictions for each user-item subgroup, our model incorporate not only topic modeling to mine item content but also social matrix factorization to handle ratings and social relationships. Experiments on an Epinions dataset show that our method significantly improves recommendation performance compared with six categories of state-of-the-art recommendation methods in terms of both prediction accuracy and recall. We have also conducted experiments to study the extent to which ratings, contexts, social relationships, and item contents contribute to recommendation performance in terms of prediction accuracy and recall.

Most controlled natural languages (CNLs) are processed with the help of a pipeline architecture that relies on different software components. We investigate in this paper in an experimental way how well answer set programming (ASP) is suited as a unifying framework for parsing a CNL, deriving a formal representation for the resulting syntax trees, and for reasoning with that representation. We start from a list of input tokens in ASP notation and show how this input can be transformed into a syntax tree using an ASP grammar and then into reified ASP rules in form of a set of facts. These facts are then processed by an ASP meta-interpreter that allows us to infer new knowledge.

Approximate dynamic programming (ADP) has proven itself in a wide range of applications spanning large-scale transportation problems, health care, revenue management, and energy systems. The design of effective ADP algorithms has many dimensions, but one crucial factor is the stepsize rule used to update a value function approximation. Many operations research applications are computationally intensive, and it is important to obtain good results quickly. Furthermore, the most popular stepsize formulas use tunable parameters and can produce very poor results if tuned improperly. We derive a new stepsize rule that optimizes the prediction error in order to improve the short-term performance of an ADP algorithm. With only one, relatively insensitive tunable parameter, the new rule adapts to the level of noise in the problem and produces faster convergence in numerical experiments.

While studying response trajectory, often the population of interest may be diverse enough to exist distinct subgroups within it and the longitudinal change in response may not be uniform in these subgroups. That is, the timeslope and/or influence of covariates in longitudinal profile may vary among these different subgroups. For example, Raudenbush (2001) used depression as an example to argue that it is incorrect to assume that all the people in a given population would be experiencing either increasing or decreasing levels of depression. In such cases, traditional linear mixed effects model (assuming common parametric form for covariates and time) is not directly applicable for the entire population as a group-averaged trajectory can mask important subgroup differences. Our aim is to identify and characterize longitudinally homogeneous subgroups based on the combination of baseline covariates in the most parsimonious way. This goal can be achieved via constructing regression tree for longitudinal data using baseline covariates as partitioning variables. We have proposed LongCART algorithm to construct regression tree for the longitudinal data. In each node, the proposed LongCART algorithm determines the need for further splitting (i.e. whether parameter(s) of longitudinal profile is influenced by any baseline attributes) via parameter instability tests and thus the decision of further splitting is type-I error controlled. We have obtained the asymptotic results for the proposed instability test and examined finite sample behavior of the whole algorithm through simulation studies. Finally, we have applied the LongCART algorithm to study the longitudinal changes in choline level among HIV patients.

We believe the relationship between these two classes of events--something not previously studied quantitatively-- could yield important insights into the nature and intrinsic dynamics of seizures. A goal of our work is to parse these complex epileptic events into distinct dynamic regimes. A challenge posed by the intracranial EEG (iEEG) data we study is the fact that the number and placement of electrodes can vary between patients. We develop a Bayesian nonparametric Markov switching process that allows for (i) shared dynamic regimes between a variable number of channels, (ii) asynchronous regime-switching, and (iii) an unknown dictionary of dynamic regimes. We encode a sparse and changing set of dependencies between the channels using a Markov-switching Gaussian graphical model for the innovations process driving the channel dynamics and demonstrate the importance of this model in parsing and out-of-sample predictions of iEEG data. We show that our model produces intuitive state assignments that can help automate clinical analysis of seizures and enable the comparison of subclinical bursts and full clinical seizures. Keywords: Bayesian nonparametric, EEG, factorial hidden Markov model, graphical model, time series 1. Introduction Despite over three decades of research, we still have very little idea of what defines a seizure.