The goal of information extraction is to extract database records from text or semi-structured sources. Traditionally, information extraction proceeds by first segmenting each candidate record separately, and then merging records that refer to the same entities. While computationally efficient, this approach is suboptimal, because it ignores the fact that segmenting one candidate record can help to segment similar ones. For example, resolving a well-segmented field with a lessclear one can disambiguate the latter’s boundaries. In this paper we propose a joint approach to information extraction, where segmentation of all records and entity resolution are performed together in a single integrated inference process. While a number of previous authors have taken steps in this direction (e.g., Pasula et al. (2003), Wellner et al. (2004)), to our knowledge this is the first fully joint approach. In experiments on the CiteSeer and Cora citation matching datasets, joint inference improved accuracy, and our approach outperformed previous ones. Further, by using Markov logic and the existing algorithms for it, our solution consisted mainly of writing the appropriate logical formulas, and required much less engineering than previous ones.

Abstract. Markov logic networks (MLNs) combine Markov networks and first-order logic, and are a powerful and increasingly popular representation for statistical relational learning. The state-of-the-art method for discriminative learning of MLN weights is the voted perceptron algorithm, which is essentially gradient descent with an MPE approximation to the expected sufficient statistics (true clause counts). Unfortunately, these can vary widely between clauses, causing the learning problem to be highly ill-conditioned, and making gradient descent very slow. In this paper, we explore several alternatives, from per-weight learning rates to second-order methods. In particular, we focus on two approaches that avoid computing the partition function: diagonal Newton and scaled conjugate gradient. In experiments on standard SRL datasets, we obtain order-of-magnitude speedups, or more accurate models given comparable learning times. 1

Programmers regularly use search as part of the development process, attempting to identify an appropriate API for a problem, seeking more information about an API, and seeking samples that show how to use an API. However, neither general-purpose search engines nor existing code search engines currently fit their needs, in large part because the information programmers need is distributed across many pages. We present Assieme, a Web search interface that effectively supports common programming search tasks by combining information from Web-accessible Java Archive (JAR) files, API documentation, and pages that include explanatory text and sample code. Assieme uses a novel approach to finding and resolving implicit references to Java packages, types, and members within sample code on the Web. In a study of programmers performing searches related to common programming tasks, we show that programmers obtain better solutions, using fewer queries, in the same amount of time spent using a general Web search interface. ACM Classification

Abstract — We present a declarative framework for collective deduplication of entity references in the presence of constraints. Constraints occur naturally in many data cleaning domains and can improve the quality of deduplication. An example of a constraint is “each paper has a unique publication venue”; iftwo paper references are duplicates, then their associated conference references must be duplicates as well. Our framework supports collective deduplication, meaning that we can dedupe both paper references and conference references collectively in the example above. Our framework is based on a simple declarative Datalogstyle language with precise semantics. Most previous work on deduplication either ignore constraints or use them in an ad-hoc domain-specific manner. We also present efficient algorithms to support the framework. Our algorithms have precise theoretical guarantees for a large subclass of our framework. We show, using a prototype implementation, that our algorithms scale to very large datasets. We provide thorough experimental results over real-world data demonstrating the utility of our framework for high-quality and scalable deduplication. I.

Markov logic networks (MLNs) combine logic and probability by attaching weights to first-order clauses, and viewing these as templates for features of Markov networks. Learning MLN structure from a relational database involves learning the clauses and weights. The state-of-the-art MLN structure learners all involve some element of greedily generating candidate clauses, and are susceptible to local optima. To address this problem, we present an approach that directly utilizes the data in constructing candidates. A relational database can be viewed as a hypergraph with constants as nodes and relations as hyperedges. We find paths of true ground atoms in the hypergraph that are connected via their arguments. To make this tractable (there are exponentially many paths in the hypergraph), we lift the hypergraph by jointly clustering the constants to form higherlevel concepts, and find paths in it. We variabilize the ground atoms in each path, and use them to form clauses, which are evaluated using a pseudolikelihood measure. In our experiments on three real-world datasets, we find that our algorithm outperforms the state-of-the-art approaches. 1.

Intelligent agents must be able to handle the complexity and uncertainty of the real world. Logical AI has focused mainly on the former, and statistical AI on the latter. Markov logic combines the two by attaching weights to first-order formulas and viewing them as templates for features of Markov networks. Inference algorithms for Markov logic draw on ideas from satisfiability, Markov chain Monte Carlo and knowledge-based model construction. Learning algorithms are based on the voted perceptron, pseudo-likelihood and inductive logic programming. Markov logic has been successfully applied to problems in entity resolution, link prediction, information extraction and others, and is the basis of the open-source Alchemy system.

We address the problem of visual event recognition in surveillance where noise and missing observations are serious problems. Common sense domain knowledge is exploited to overcome them. The knowledge is represented as first-order logic production rules with associated weights to indicate their confidence. These rules are used in combination with a relaxed deduction algorithm to construct a network of grounded atoms, the Markov Logic Network. The network is used to perform probabilistic inference for input queries about events of interest. The system’s performance is demonstrated on a number of videos from a parking lot domain that contains complex interactions of people and vehicles.

The task of identifying synonymous relations and objects, or synonym resolution, is critical for high-quality information extraction. This paper investigates synonym resolution in the context of unsupervised information extraction, where neither hand-tagged training examples nor domain knowledge is available. The paper presents a scalable, fullyimplemented system that runs in O(KN log N) time in the number of extractions, N, and the maximum number of synonyms per word, K. The system, called Resolver, introduces a probabilistic relational model for predicting whether two strings are co-referential based on the similarity of the assertions containing them. On a set of two million assertions extracted from the Web, Resolver resolves objects with 78 % precision and 68 % recall, and resolves relations with 90 % precision and 35 % recall. Several variations of Resolver’s probabilistic model are explored, and experiments demonstrate that under appropriate conditions these variations can improve F1 by 5%. An extension to the basic Resolver system allows it to handle polysemous names with 97 % precision and 95 % recall on a data set from the TREC corpus.

Many real-world applications of AI require both probability and first-order logic to deal with uncertainty and structural complexity. Logical AI has focused mainly on handling complexity, and statistical AI on handling uncertainty. Markov Logic Networks (MLNs) are a powerful representation that combine Markov Networks (MNs) and first-order logic by attaching weights to first-order formulas and viewing these as templates for features of MNs. State-of-theart structure learning algorithms of MLNs maximize the likelihood of a relational database by performing a greedy search in the space of candidates. This can lead to suboptimal results because of the incapability of these approaches to escape local optima. Moreover, due to the combinatorially explosive space of potential candidates these methods are computationally prohibitive. We propose a novel algorithm for learning MLNs structure, based on the Iterated Local Search (ILS) metaheuristic that explores the space of structures through a biased sampling of the set of local optima. The algorithm focuses the search not on the full space of solutions but on a smaller subspace defined by the solutions that are locally optimal for the optimization engine. We show through experiments in two real-world domains that the proposed approach improves accuracy and learning time over the existing state-of-the-art algorithms. 1

The task of identifying synonymous relations and objects, or Synonym Resolution (SR), is critical for high-quality information extraction. The bulk of previous SR work assumed strong domain knowledge or hand-tagged training examples. This paper investigates SR in the context of unsupervised information extraction, where neither is available. The paper presents a scalable, fully-implemented system for SR that runs in O(KN log N) time in the number of extractions N and the maximum number of synonyms per word, K. The system, called RESOLVER, introduces a probabilistic relational model for predicting whether two strings are co-referential based on the similarity of the assertions containing them. Given two million assertions extracted from the Web, RESOLVER resolves objects with 78 % precision and an estimated 68 % recall and resolves relations with 90 % precision and 35 % recall. 1