An algorithm is described in this paper for the automatic resolution of intra-sentential pronominal anaphoric references in Chinese sentences. 

  Anaphora is cohesion (presupposition) which points back to some previous item. Anaphora resolution is a complicated problem in Natural Language Processing and has attracted the attention of many researchers. 

  The method is based on sentences category (SC) and makes use of its representation formula. SC is the main concept of HNC theory -- Hierarchical Network of Concepts theory, which is a novel theory of natural language processing. In our approach, some personal noun phrases (NPs) preceding an anaphor are initially regarded as potential candidates for antecedents. Then, their semantic role are compared with that of personal pronoun, relying on a set of anaphora resolution factors. These factors can be "eliminating", i.e. discounting certain noun phrases from the set of possible candidates (such as gender and number constraints) or "preferential", giving more preference to certain candidates and less to others. All these factors are presented in this paper. The ability to perform anaphora resolution is important in speech recognition. Our algorithm is used to enhance the recognition ability of IBM Via Voice.

Key words: natural language processing / anaphora resolution / HNC theory / speech recognition