Summarization

Read less. Learn more.

Summarization systems - like Language Computer's GistTexter - can reduce the time it takes a user to synthesize information from a collection of documents by producing short, targeted summaries that can be read and interpreted quickly.

Automatic summarization allows users to distill the information content of a potentially very large set of documents into a much smaller set of passages that can be read in a matter of minutes. By analyzing the semantic representations in a topic and emphasizing the differences across multiple documents, useful abstracts can be produced for virtually any topic of interest.

Responsive Summarization

LCC's GistTexter incorporates a multi-document summarization system, which summarizes collections of documents on the same topic, and a question-focused summarization system, which produces passage-length answers to questions based on information found in the collection of documents.

  • Multi-Document Summarization: Multi-document summarization systems can be used to distill the content of hundreds of document on the same topic into a set of passages that contains only the most important information from a document collection. In olrder to produce summaries that are information and coherent, multi-document summarization systems need to be able to identify a set of sentences that most closely correspond to the topic of the document collection as a whole.
  • Question-Focused Summarization: Question-focused summarization systems provide answers to complex questions by combining results from its topic representation and sentence ranking modules with question processing technology as found in LCC's Ferret question answering system. In order to identify the information need of complex questions, GistTexter submits questions to an automatic semantic question decomposition module, which seeks to identify the set of simpler questions that can be inferred from the complex question.

LCC is committed to developing robust summarization technology. Our multi-document summarization system has been recognized to produce the most responsive summaries at the past 5 annual Document Understanding Conference (DUC) evaluations.

Related Products

GistTexter

Multi-document summarization for search results, document collections, and more.

 
Ferret

High-performance question answering for English, Arabic, Chinese, Farsi, and Korean.

 

For More Information

For more information on Language Computer's summarization technology, contact us at (972) 231-0052 or e-mail us today.