Identifying part-of-speech patterns for automatic tagging

Some part-of-speech tagging errors are very damaging to the ability to further process the text. For systems that use part-of-speech tagging as a prelude to parsing and knowledge extraction, it is imperative to have the cleanest possible tagging. A state-of-the-art rule-based tagger has an error rate of approximately 39% when annotating main verbs that have not been previously seen. We apply neural networks to this real-world problem of identifying part-of-speech patterns that indicate a main verb so as to correct the output of the rule-based tagger.