Rapidly Building High-Performance Information Agents

Abstract

This research project addressed the problem of rapidly constructing high-performance information agents. In this work we made three significant advances. First, we developed a technique called Co-testing to address the problem of generating high accuracy wrappers to extract data from online sources. Second, we developed a method called speculation execution for improving the performance of streaming data flow execution systems used in information agents. Third, we developed a wizard-based approach to interactively and rapidly constructing information agents.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 31, 2004
Accession Number
ADA427498

Entities

People

  • Craig Knoblock

Organizations

  • University of Southern California

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Artificial Intelligence
  • Automata Theory
  • Case Studies
  • Computer Languages
  • Computer Science
  • Computers
  • Data Integration
  • Data Mining
  • Information Science
  • Information Systems
  • Machine Learning
  • Network Science
  • Semi-Supervised Learning
  • Web Service
  • World Wide Web

Fields of Study

  • Computer science

Readers

  • Distributed Systems and Data Platform Development
  • Parallel and Distributed Computing.