This paper surveys three research directions in parsing. First, we look at methods for both automatically generating a set of diverse parsers and combining the outputs of different parsers into a single parse. Next, we will discuss a parsing method known as transformation-based parsing. This method, though less accurate than the best current corpus-derived parsers, is able to parse quite accurately while learning only a small set of easily understood rules, as opposed to the many-megabyte parameter files learned by other techniques. Finally, we review a recent study exploring how people and machines compare at the task of creating a program to automatically annotate noun phrases.

In this paper we describe a new technique for parsing free text: a transformational grammar is automatically learned that is capable of accurately parsing text into binary-branching syntactic trees. The algorithm works by beginning in a very naive state of knowledge about phrase structure. By repeatedly comparing the results of bracketing in the current state to proper bracketing provided in the training corpus, the system learns a set of simple structural transformations that can be applied to reduce the number of errors. After describing the algorithm, we present results and compare these results to other recent results in automatic grammar induction.