A Method of Incorporating Bigram Constraints into an LR Table and Its E ectiveness in Natural Language Processing

In this paper, we propose a method for constructing bigram LR tables by way of incorporating bigram constraints into an LR table. Using a bigram LR table, it is possible for a GLR parser to make use of both bigram and CFG constraints in natural language processing. Applying bigram LR tables to our GLR method has the following advantages: (1) Language models utilizing bigram LR tables have lower perplexity than simple bigram language models, since local constraints (bigram) and global constraints (CFG) are combined in a single bigram LR table. (2) Bigram constraints are easily acquired from a given corpus. Therefore data sparseness is not likely to arise. (3) Separation of local and global constraints keeps down the number of CFG rules. The rst advantage leads to a reduction in complexity, and as the result, better performance in GLR parsing. Our experiments demonstrate the e ectiveness of our method.