GLR Parser with Conditional Action Model(CAM)

There are two different approaches in the LR parsing. The first one is the deterministic approach that performs the only one action using the control rules learned without any LR parsing resource. It shows good performance in speed. But it has a disadvantage that it cannot correct the previous mistakes, thus directly affects the parsing result. The second one is the probabilistic LR parsing approach, which uses annotated corpora and LR table. In this paper, we propose a new probabilistic GLR parsing method that can solve the problems of conventional LR parsing approaches. The parser takes probabilistic shift or reduce action on the graph-structured stack, which is conditioned on the partially constructed parse. It chooses the most probable action sequences for the given input sentence when it reaches the final state. With the probabilistic model, it can overcome the shortcomings of the deterministic LR parsing scheme, and it uses richer information in the stack than the previous probabilistic methods.