WSJ (via Tech Review) reports on research which uses “gene comparison techniques to identify word patterns from different news sources describing the same event.”
The scheme could eventually allow computers to more easily process natural language, produce paraphrases that could be used in machine translation, and help people who have trouble reading certain types of sentences.
Given a sentence to paraphrase, the system finds the closest match among one set of lattices, then uses the matching lattice from the other source to fill in the argument values of the original sentence. The prototype can produce as many as 20 paraphrases per sentence, according to the researchers.
TRN Magazine has more. One of the researchers is Regina Barzilay.