Communications of the ACM
Mining For Meaning: Getting Computers to nderstand Natural Language Texts
A new procedure will allow computers to learn to identify semantically relevant relationships within texts.
Credit: CockingASnook.wordpress.com
Computer scientists at Saarland University and the University of Amsterdam are developing a procedure that will enable computers to learn to identify semantically relevant relationships within texts.
Computers that are able to understand natural language texts could potentially analyze texts and provide specific answers about them. "The model that we have developed simulates how humans create texts," says Saarland's Ivan Titov. "In order to understand texts, we get our computers to work through this process but in the reverse direction: given the text the computer will uncover its meaning or even intent of the writer."
The team uses millions of sentences to generate both the model and the rules contained within it. The analysis of the massive dataset requires a lot of computing power, with specially developed algorithms running on nearly 100 computers.
The researchers eventually hope to use the method to get machines to automatically summarize short texts and answer questions about the content.
From Saarland University
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found