Inference in large scale graphical models is an important task in many domains, and in particular probabilistic relational models (e.g. Markov logic networks). Such models often exhibit considerable symmetry, and it is a challenge to devise algorithms that exploit this symmetry to speed up inference. Recently, the automorphism group has been proposed to formalize mathematically what "exploiting symmetry" means. However, obtaining symmetry derived from automorphism is GI-hard, and consequently only a small fraction of the symmetry is easily available for effective employment. In this paper, we improve upon efficiency in two ways. First, we introduce the Cluster Signature Graph (CSG), a platform on which greater portions of the symmetries can be revealed and exploited. CSGs classify clusters of variables by projecting relations between cluster members onto a graph, allowing for the efficient pruning of symmetrical clusters even before their generation. Second, we introduce a novel framework based on CSGs for the Sherali-Adams hierarchy of linear program (LP) relaxations, dedicated to exploiting this symmetry for the benefit of tight Maximum A Posteriori (MAP) approximations. Combined with the pruning power of CSG, the framework quickly generates compact formulations for otherwise intractable LPs, as demonstrated by several empirical results.

Mutual information (MI) based approaches are a popular feature selection paradigm. Although the stated goal of MI-based feature selection is to identify a subset of features that share the highest mutual information with the class variable, most current MI-based techniques are greedy methods that make use of low dimensional MI quantities. The reason for using low dimensional approximation has been mostly attributed to the difficulty associated with estimating the high dimensional MI from limited samples. In this paper, we argue a different viewpoint that, given a very large amount of data, the high dimensional MI objective is still problematic to be employed as a meaningful optimization criterion, due to its overfitting nature: the MI almost always increases as more features are added, thus leading to a trivial solution which includes all features. We propose a novel approach to the MI-based feature selection problem, in which the overfitting phenomenon is controlled rigourously by means of a statistical test. We develop local and global optimization algorithms for this new feature selection model, and demonstrate its effectiveness in the applications of explaining variables and objects.

In situated dialogue with artificial agents (e.g., robots), although a human and an agent are co-present, the agent's representation and the human's representation of the shared environment are significantly mismatched. Because of this misalignment, our previous work has shown that when the agent applies traditional approaches to generate referring expressions for describing target objects with minimum descriptions, the intended objects often cannot be correctly identified by the human. To address this problem, motivated by collaborative behaviors in human referential communication, we have developed two collaborative models - an episodic model and an installment model - for referring expression generation. Both models, instead of generating a single referring expression to describe a target object as in the previous work, generate multiple small expressions that lead to the target object with the goal of minimizing the collaborative effort. In particular, our installment model incorporates human feedback in a reinforcement learning framework to learn the optimal generation strategies. Our empirical results have shown that the episodic model and the installment model outperform previous non-collaborative models with an absolute gain of 6% and 21% respectively.

Humorous images can be seen in many social media websites. However, newcomers to these websites often have trouble fitting in because of the subculture among the community is usually implicit. Among all the types of humorous images, Internet memes are relatively hard for newcomers to understand. In this work, we develop a system leveraging crowdsourcing technique to generate explanations for meme images. We claim that people who are not familiar with Internet meme subculture can still quickly pick up the gist of the memes through reading the explanations. Our template-based explanation can illustrate the incongruity between normal situations and the punchlines in jokes. The explanations can be produced by going through 2 designed humor tasks. In our pilot study, acceptable explanations for 5 unique memes are generated. For further study, generating explanations for more general text jokes are possible.

Our research aims at building interactive robots and agents that can expand their knowledge by interacting with human users. In this paper, we focus on learning goal-oriented tasks from situated interactive instructions. Learning the structure of novel tasks and how to execute them is a challenging computational problem requiring the agent to acquire a variety of knowledge including goal definitions and hierarchical control information. We frame acquisition of novel tasks as an explanation-based learning (EBL) problem and propose an interactive learning variant of EBL for a robotic agent. We show that our approach can exploit information in situated instructions along with the domain knowledge to demonstrate fast generalization on several tasks. The knowledge acquired transfers across structurally similar tasks. Finally, we show that our approach seamlessly combines agent-driven exploration with instructions for mixed-initiative learning.

Low-rank coding (LRC), originated from matrix decomposition, is recently introduced into image classification. Following the standard bag-of-words (BOW) pipeline, when coding the data matrix in the sense of low-rankness incorporates contextual information into the traditional BOW model, this can capture the dependency relationship among neighbor patches. It differs from the traditional sparse coding paradigms which encode patches independently. Current LRC-based methods use l_1 norm to increase the discrimination and sparseness of the learned codes. However, such methods fail to consider the local manifold structure between dataspace and dictionary space. To solve this problem, we propose a locality-constrained low-rank coding (LCLR) algorithm for image representations. By using the geometric structure information as a regularization term,we can obtain more discriminative representations. In addition, we present a fast and stable online algorithmto solve the optimization problem. In the experiments,we evaluate LCLR with four benchmarks, including one face recognition dataset (extended Yale B), one handwrittendigit recognition dataset (USPS), and two image datasets (Scene13 for scene recognition and Caltech101 for object recognition). Experimental results show thatour approach outperforms many state-of-the-art algorithmseven with a linear classifier.

Multiple Instance Learning (MIL) is a popular learning technique in various vision tasks including image classification. However, most existing MIL methods do not consider the problem of insufficient examples in the given target category. In this case, it is difficult for traditional MIL methods to build an accurate classifier due to the lack of training examples. Motivated by the empirical success of transfer learning, this paper proposes a novel approach of Adaptive Knowledge Transfer for Multiple Instance Learning (AKT-MIL) in image classification. The new method transfers cross-category knowledge from source categories under multiple instance setting for boosting the learning process. A unified learning framework with a data-dependent mixture model is designed to adaptively combine the transferred knowledge from sources with a weak classifier built in the target domain. Based on this framework, an iterative coordinate descent method with Constraint Concave-Convex Programming (CCCP) is proposed as the optimization procedure. An extensive set of experimental results demonstrate that the proposed AKT-MIL approach substantially outperforms several state-of-the-art algorithms on two benchmark datasets, especially in the scenario when very few training examples are available in the target domain.

High profile large scale public events are attractive targets for terrorist attacks. The recent Boston Marathon bombings on April 15, 2013 have further emphasized the importance of protecting public events. The security challenge is exacerbated by the dynamic nature of such events: e.g., the impact of an attack at different locations changes over time as the Boston marathon participants and spectators move along the race track. In addition, the defender can relocate security resources among potential attack targets at any time and the attacker may act at any time during the event. This paper focuses on developing efficient patrolling algorithms for such dynamic domains with continuous strategy spaces for both the defender and the attacker. We aim at computing optimal pure defender strategies, since an attacker does not have an opportunity to learn and respond to mixed strategies due to the relative infrequency of such events. We propose SCOUT-A, which makes assumptions on relocation cost, exploits payoff representation and computes optimal solutions efficiently. We also propose SCOUT-C to compute the exact optimal defender strategy for general cases despite the continuous strategy spaces. SCOUT-C computes the optimal defender strategy by constructing an equivalent game with discrete defender strategy space, then solving the constructed game. Experimental results show that both SCOUT-A and SCOUT-C significantly outperform other existing strategies.

Markov processes are widely used to generate sequences that imitate a given style, using random walk. Random walk generates sequences by iteratively concatenating states to prefixes of length equal or less than the given Markov order}. However, at higher orders, Markov chains tend to replicate chunks of the corpus with a size possibly higher than the order, a primary form of plagiarism. The Markov order defines a maximum length for training but not for generation. In the framework of constraint satisfaction (CSP), we introduce MaxOrder. This global constraint ensures that generated sequences do not include chunks larger than a given maximum order. We exhibit an automaton that recognises the solution set, with a size linear in the size of the corpus. We propose a linear-time procedure to generate this automaton from a corpus and a given max order. We then use this automaton to achieve generalised arc consistency for the MaxOrder constraint, holding on a sequence of size n, in O(n.T) time, where T is the size of the automaton. We illustrate our approach by generating text sequences from text corpora with a maximum order guarantee, effectively controlling plagiarism.

Tweets ranking is important for information acquisition in Microblog. Due to the content sparsity and lackof labeled data, it is better to employ semi-supervisedlearning methods to utilize the unlabeled data. However,most of previous semi-supervised learning methods donot consider the pair conflict problem, which means thatthe new selected unlabeled data may conflict with the labeled and previously selected data. It will hurt the learning performance a lot, if the training data contains manyconflict pairs. In this paper, we propose a new collaborative semi-supervised SVM ranking model (CSR-TC)with consideration of the order conflict. The unlabeleddata is selected based on a dynamically maintained transitive closure graph to avoid pair conflict. We also investigate the two views of features, intrinsic and contentrelevant features, for the proposed model. Extensive experiments are conducted on TREC Microblogging corpus. The results demonstrate that our proposed methodachieves significant improvement, compared to severalstate-of-the-art models.