This paper presents a two-step approach to determining whether a transliterated personal name from different Chinese texts stands for the same referent. A heuristic strategy based on biographical information and “colleague” names is firstly used to form an initial set of coreference chains, and then, a clustering algorithm based Vector Space Model (VSM) is applied to merge chains under the control of a full name consistent constraint. Experimental results show that this approach achieves a good performance.