A Theory-Refinement Approach to Information Extraction

We investigate applying theory renemen t to the task of extracting information from text. In theory renemen t, partial domain knowledge (which may be incorrect) is given to a supervised learner. The provided knowledge guides the learner in its task, but the learner can rene or even discard this knowledge during training. Our supervised learner is a \knowledge-based" neural network that initially contains \compiled" prior knowledge about a particular information extraction (IE) task. The prior knowledge needs to specify the extraction slots for the specic IE task. Our approach uses generate-and-test to address the IE task. In the generation step, we produce candidate extractions by intelligently searching the space of possible extractions. In the test step, we use the trained network to judge each candidate and output those that exceed a system-selected threshold. Experiments on the CMU seminarannouncements and the Yeast subcellularlocalization domains demonstrate our approach’s value.