While parallelizing a program, to ensure the correctness of the result, a parallel compiler can only adopt a conservative pol icy that if it cannot accurately determine whether two pieces of code wi ll conflict while being executed in parallel, it does not permit them to be executed in parallel. Although this approach can ensure correctness, it also limits the amount of parallelism that can be exploited. A number of speculative multithreading (SpMT) approaches gained wide popularity because they can get more parallelism by allowing two pieces of potentially conflicting code to execute in parallel while ensuring the correctness of the results by recovering from conflicts. However, the dividing-along-the-control-flow-path method employed by the traditional SpMT approach is unsuitable for a number of programs in which operations on different data structures are interleaved in the control flow path. If the program is linearly divided into threads along the control flow path, the operations on the same data structure will be divided among the different threads that are doomed to be prone to conflict when being executed concurrently. To effectively parallelize these programs, this paper proposes a data structure centric approach, in which the objects of a program are organized into different groups and the operations on the objects wit hin the same group are dispatched to the same thread for execution, thereby reducing the likelihood of conflict on the same data structure.