The role of documents vs. queries in extracting class attributes from text
Published 6 November 2007
Marius Paşca, Benjamin Van Durme, Nikesh Garera
Citations26
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
In a head-to-head qualitative comparison, a lightweight extraction method produces class attributes that are 45% more accurate on average, when acquired from query logs rather than Web documents, as part of a large-scale study on extracting prominent attributes or quantifiable properties from unstructured text.
Abstract
Challenging the implicit reliance on document collections, this paper discusses the pros and cons of using query logs rather than document collections, as self-contained sources of data in textual information extraction. The differences are quantified as part of a large-scale study on extracting prominent attributes or quantifiable properties of classes (e.g., top speed, price and fuel consumption for CarModel) from unstructured text. In a head-to-head qualitative comparison, a lightweight extraction method produces class attributes that are 45% more accurate on average, when acquired from query logs rather than Web documents.
Keywords
Computer Science
Learning question classifiers
1,276 Citations2002Xin Li, Dan Roth
A hierarchical classifier is learned that is guided by a layered semantic hierarchy of answer types, and eventually classifies questions into fine-grained classes.
<i>Snowball</i>
1,169 Citations2000Eugene Agichtein, Luis Gravano
This paper develops a scalable evaluation methodology and metrics for the task, and presents a thorough experimental evaluation of Snowball and comparable techniques over a collection of more than 300,000 newspaper documents.
Improving web search ranking by incorporating user behavior information
1,041 Citations2006Eugene Agichtein, Eric Brill +1 more
It is shown that incorporating user behavior data can significantly improve ordering of top results in real web search setting, improving the accuracy of a competitive web search ranking algorithms by as much as 31% relative to the original performance.
Studies in linguistics and philosophyIntroduction to Montague Semantics
652 Citations1980David R. Dowty, Robert E. Wall +1 more
This book discusses Montague's Intensional Logic, a Higher-Order Type-Theoretic Language, and some Unresolved Issues with Possible Worlds Semantics and Propositional Attitudes.
Espresso
535 Citations2006Patrick Pantel, Marco Pennacchiotti
Experimental results show that the exploitation of generic patterns substantially increases system recall with small effect on overall precision, and a principled measure of pattern and instance reliability enabling the filtering algorithm.
Probabilistic query expansion using query logs
424 Citations2002Hang Cui, Ji-Rong Wen +2 more
The experimental results show that the log-based probabilistic query expansion method can greatly improve the search performance and has several advantages over other existing methods.
ACM SIGKDD Explorations NewsletterMining knowledge from text using information extraction
296 Citations2005Raymond J. Mooney, Răzvan Bunescu
Methods and implemented systems for information extraction systems used to extract concrete data from a set of documents are discussed and results on mining real text corpora of biomedical abstracts, job announcements, and product descriptions are summarized.
Preemptive information extraction using unrestricted relation discovery
240 Citations2006Yusuke Shinyama, Satoshi Sekine
A technique called Unrestricted Relation Discovery is proposed that discovers all possible relations from texts and presents them as tables in order to extend the boundary of Information Extraction systems.
North American Chapter of the Association for Computational LinguisticsAutomatically Labeling Semantic Classes
220 Citations2004Patrick Pantel, Deepak Ravichandran
An algorithm is proposed for labeling semantic classes and for leveraging them to extract is-a relationships using a top-down approach to address the limitations of broad-coverage lexical resources such as WordNet and Cyc.
Concept discovery from text
185 Citations2002Dekang Lin, Patrick Pantel
A clustering algorithm called CBC (Clustering By Committee) that automatically discovers concepts from text that outperforms several well-known clustering algorithms in cluster quality.
International Joint Conference on Artificial IntelligenceWhat you seek is what you get: extraction of class attributes from query logs
128 Citations2007Marius Paşca, Benjamin Van Durme
The role of Web query logs, rather than Web documents, are explored as an alternative source of class attributes, and the quality of the extracted attributes recommends query logs as a valuable, albeit little explored, resource for information extraction.
Exploring distributional similarity based models for query spelling correction
116 Citations2006Mu Li, Yang Zhang +2 more
Novel methods for use of distributional similarity estimated from query logs in learning improved query spelling correction models are described, which can significantly outperform their baseline systems in the web query spelling Correction task.
KnowItNow
106 Citations2005Michael Cafarella, Doug Downey +2 more
A novel architecture for IE that obviates queries to commercial search engines is introduced, embodied in a system called KnowItNow that performs high-precision IE in minutes instead of days, and the tradeoff between recall and speed is quantified.
Acquiring Hyponymy Relations from Web Documents.
78 Citations2004Keiji Shinzato, Kentaro Torisawa
This paper describes an automatic method for acquiring hyponymy relations from HTML documents on the WWW that uses clues such as itemization or listing in HTML documents and statistical measures such as document frequencies and verb-noun co-occurrences.
Lecture notes in computer scienceAutomatic Discovery of Attribute Words from Web Documents
50 Citations2005Kosuke Tokunaga, Jun’ichi Kazama +1 more
It is confirmed that Web.com is no longer in danger of beingDoS-proof.
Evaluating answers to definition questions
47 Citations2003Ellen M. Voorhees
An initial evaluation of systems that answer questions seeking definitions suggests that humans agree sufficiently as to what the basic concepts that should be included in the definition of a particular subject are to permit the computation of concept recall.
Re-ranking search results using query logs
37 Citations2006Ziming Zhuang, Silviu Cucerzan
Evaluation results show that Q-Rank gains a considerable advantage over the current ranking system of a large-scale commercial Web search engine, being able to improve the relevance of search results for 82% of the queries.
An analysis of knowledge collected from volunteer contributors
34 Citations2005Timothy Chklovski, Yolanda Gil
Analysis of a representative collection of part-of information contributed by volunteers reveals that redundancy of contribution helps identify high quality statements, but that some of the statements also have overly high redundancy, drawing contributor effort away from areas where they are needed more.
Research Showcase @ Carnegie Mellon University (Carnegie Mellon University)Instance-Based Question Answering: A Data-Driven Approach
32 Citations2018Lucian Vlad Lita, Jaime Carbonell
This work proposes a principled, datadriven Instance-Based approach to Question Answering that treats answer extraction as a binary classification problem in which text snippets are labeled as correct or incorrect answers.
National Conference on Artificial IntelligenceTuring's dream and the knowledge challenge
23 Citations2006Lenhart K. Schubert
There is a set of clear-cut challenges, all centering around knowledge, that have received insufficient attention in AI and whose solution could bring the realization of Turing's dream of a machine the authors can talk with just like a person, and which is therefore (at least) their intellectual equal.
Extracting key semantic terms from Chinese speech query for web searches
5 Citations2003Gang Wang, Tat‐Seng Chua +1 more
The main contribution of this research is in devising a divide-and-conquer strategy to alleviate the speech recognition errors by using the query model to facilitate the extraction of main core semantic string (CSS) from the Chinese natural language speech query.