Implicitly structured content on the Web such as HTML tables and lists can be extremely valuable for web search, question answering, and information retrieval, as the implicit structure in a page often reflects the underlying semantics of the data. Unfortunately, exploiting this information presents significant challenges due to the immense amount of implicitly structured content on the web, lack of schema information, and unknown source quality. We present TQA, a web-scale system for automatic question answering that is often able to find answers to real natural language questions from the implicitly structured content on the web. Our experiments over more than 200 million structures extracted from a partial web crawl demonstrate the promise of our approach.

Many applications in text processing require significant human effort for either labeling large document collections (when learning statistical models) or extrapolating rules from them (when using knowledge engineering). In this work, we describe a way to reduce this effort, while retaining the methods ’ accuracy, by constructing a hybrid classifier that utilizes human reasoning over automatically discovered text patterns to complement machine learning. Using a standard sentiment-classification dataset and real customer feedback data, we demonstrate that the resulting technique results in significant reduction of the human effort required to obtain a given classification accuracy. Moreover, the hybrid text classifier also results in a significant boost in accuracy over machine-learning based classifiers when a comparable amount of labeled data is used.

We show that incorporating user behavior data can significantly improve ordering of top results in real web search setting. We examine alternatives for incorporating feedback into the ranking process and explore the contributions of user feedback compared to other common web search features. We report results of a large scale evaluation over 3,000 queries and 12 million user interactions with a popular web search engine. We show that incorporating implicit feedback can augment other features, improving the accuracy of a competitive web search ranking algorithms by as much as 31% relative to the original performance.

Evaluating user preferences of web search results is crucial for search engine development, deployment, and maintenance. We present a real-world study of modeling the behavior of web search users to predict web search result preferences. Accurate modeling and interpretation of user behavior has important applications to ranking, click spam detection, web search personalization, and other tasks. Our key insight to improving robustness of interpreting implicit feedback is to model query-dependent deviations from the expected "noisy" user behavior. We show that our model of clickthrough interpretation improves prediction accuracy over state-of-the-art clickthrough methods. We generalize our approach to model user behavior beyond clickthrough, which results in higher preference prediction accuracy than models based on clickthrough information alone. We report results of a large-scale experimental evaluation that show substantial improvements over published implicit feedback interpretation methods.

We study the effect of user supplied relevance feedback in improving web search results. Rather than using query refinement or document similarity measures to rerank results, we show that the web-graph distance between two documents is a robust measure of their relative relevancy. We demonstrate how the use of this metric can improve the rankings of result URLs, even when the user only rates one document in the dataset. Our research suggests that such interactive systems can significantly improve search results.

Since the publication of Brin and Page's paper on PageRank, many in the Web community have depended on PageRank for the static (query-independent) ordering of Web pages. We show that we can significantly outperform PageRank using features that are independent of the link structure of the Web. We gain a further boost in accuracy by using data on the frequency at which users visit Web pages. We use RankNet, a ranking machine learning algorithm, to combine these and other static features based on anchor text and domain characteristics. The resulting model achieves a static ranking pairwise accuracy of 67.3% (vs. 56.7% for PageRank or 50% for random).

We present an analysis of the structured relationships observed in a randomly sampled set of question-like queries submitted to a search engine for a popular online encyclopedic document collection. Our study shows that a relatively small number of binary relationships account for most of the queries in the sample. This empirically validates an approach of analyzing query logs to identify the relationships most relevant to user needs and populating corresponding fact tables from the collection for factoid question answering. Our analysis shows that such an approach can lead to substantial coverage of user questions.

In this paper we propose a unified framework for automatic evaluation of NLP applications using N-gram co-occurrence statistics. The automatic evaluation metrics proposed to date for Machine Translation and Automatic Summarization are particular instances from the family of metrics we propose. We show that different members of the same family of metrics explain best the variations obtained with human evaluations, according to the application being evaluated (Machine Translation, Automatic Summarization, and Automatic Question Answering) and the evaluation guidelines used by humans for evaluating such applications.

We describe a method to define and use subwebs, user-defined neighborhoods of the Internet. Subwebs help improve search performance by inducing a topic-specific page relevance bias over a collection of documents. Subwebs may be automatically identified using a simple algorithm we describe, and used to provide highly-relevant topic-specific information retrieval. Using subwebs in a Help and Support topic, we see marked improvements in precision compared to generic search engine results.

Web communities are web virtual broadcasting spaces where people can freely discuss anything. While such communities function as discussion boards, they have even greater value as large repositories of archived information. In order to unlock the value of this resource, we need an effective means for searching archived discussion threads. Unfortunately the techniques that have proven successful for searching document collections and the Web are not ideally suited to the task of searching archived community discussions. In this paper, we explore the problem of creating an effective ranking function to predict the most relevant messages to queries in community search. We extract a set of predictive features from the thread trees of newsgroup messages as well as features of message authors and lexical distribution within a message thread. Our final results indicate that when using linear regression with this feature set, our search system achieved a 28.5% performance improvement compared to our baseline system.

We describe the architecture of the AskMSR question answering system and systematically evaluate contributions of different system components to accuracy. The system differs from most question answering systems in its dependency on data redundancy rather than sophisticated linguistic analyses of either questions or candidate answers. Because a wrong answer is often worse than no answer, we also explore strategies for predicting when the question answering system is likely to give an incorrect answer.

This paper describes a question answering system that is designed to capitalize on the tremendous amount of data that is now available online. Most question answering systems use a wide variety of linguistic resources. We focus instead on the redundancy available in large corpora as an important resource. We use this redundancy to simplify the query rewrites that we need to use, and to support answer mining from returned snippets. Our system performs quite well given the simplicity of the techniques being utilized. Experimental results show that question answering accuracy can be greatly improved by analyzing more and more matching passages. Simple passage ranking and n-gram extraction techniques work well in our system making it efficient to use with many backend retrieval engines.

The amount of readily available on-line text has reached hundreds of billions of words and continues to grow. Yet for most core natural language tasks, algorithms continue to be optimized, tested and compared after training on corpora consisting of only one million words or less. In this paper, we evaluate the performance of different learning methods on a prototypical natural language disambiguation task, confusion set disambiguation, when trained on orders of magnitude more labeled data than has previously been used. We are fortunate that for this particular application, correctly labeled training data is free. Since this will often not be the case, we examine methods for effectively exploiting very large corpora when labeled data comes at a cost.

The noisy channel model has been applied to a wide range of problems, including spelling correction. These models consist of two components: a source model and a channel model. Very little research has gone into improving the channel model for spelling correction. This paper describes a new channel model for spelling correction, based on generic string to string edits. Using this model gives significant performance improvements compared to previously proposed models.

A wide range of natural language problems can be viewed as disambiguating between a small set of alternatives based upon the string context surrounding the ambiguity site. In this paper we demonstrate that classification accuracy can be improved by invoking a more descriptive feature set than what is typically used. We present a technique that disambiguates by learning regular expressions describing the string contexts in which the ambiguity sites appear.

A great deal of work has been done demonstrating the ability of machine learning algorithms to automatically extract linguistic knowledge from annotated corpora.Very little work has gone into quantifying the difference in ability at this task between a person and a machine.This paper is a first step in that direction.

We describe a new algorithm for finding the hypothesis in a recognition lattice that is expected to minimize the word error rate (WER). Our approach thus overcomes the mismatch between the word-based performance metric and the standard MAP scoring paradigm that is sentence-based, and that can lead to sub-optimal recognition results. To this end we first find a complete alignment of all words in the recognition lattice, identifying mutually supporting and competing word hypotheses. Finally, a new sentence hypothesis is formed by concatenating the words with maximal posterior probabilities. Experimentally, this approach leads to a significant WER reduction in a large vocabulary recognition task. 1. INTRODUCTION Word lattices are used by most speech recognizers as a compact representation of a set of alternative hypotheses. In the standard MAP decoding approach [1] the recognizer outputs the string of words corresponding to the path with the highest posterior probability given the acous...

In order to continue building systems with progressively more complex natural language capabilities, it is crucial that great strides are made toward solving the core linguistic analysis problems for complex and possibly unrestricted domains. A great deal of progress has been made by applying machine learning techniques to automatically train systems from manually annotated corpora to provide detailed linguistic analyses to sentences. This paper examines a number of issues within this paradigm of automatic linguistic knowledge acquisition and how they relate to pushing progress in the field of natural language processing over the next decade.

In transformation-based parsing, a finite sequence of tree rewriting rules are checked for application to an input structure. Since in practice only a small percentage of rules are applied to any particular structure, the naive parsing algorithm is rather inefficient. We exploit this sparseness in rule applications to derive an algorithm two to three orders of magnitude faster than the standard parsing algorithm.

In this paper we describe a new technique for parsing free text: a transformational grammar1 is automatically learned that is capable of accurately parsing text into binary-branching syntactic trees with nonterminals unlabelled. The algorithm works by beginning in a very naive state of knowledge about phrase structure. By repeatedly comparing the results of bracketing in the current state to proper bracketing provided in the training corpus, the system learns a set of simple structural transformations that can be applied to reduce error. After describing the algorithm, we present results and compare these results to other recent results in automatic grammar induction.

No abstract available.