基于条件随机场的分词工作流研究与应用

Research and Application of Word Segmentation Workflow Based on Conditional Random Field

为了提高分词处理的执行效率,将工作流引入到具体的分词处理中,设计实现了基于条件随机场的分词工作流应用。首先分析了条件随机场分词处理过程,构造了分词工作流应用,给出了工作流描述文档的部分片段,并就部分成员服务如何调用自调度算法进行了分析,然后详细介绍了工作流成员服务的实现方法,并给出了关键代码。实验结果表明该模型不仅有效地缩短了分词处理程序的执行时间,而且能够提高网格资源的利用率。

In order to improve the implementation efficiency of word segmentation processing, the paper intoduces workflow into specific word segmentation processing, designs and realizes word segmentation workflow application based on conditional random field. First analyze the word segmentation processing based on conditional random field, construct word segmentation workflow application, gives the fragment of workflow description documents and analyzes some member services to how to call self - scheduling algorithm, then in detail introduces implementation method of workiqow member services, and gives some key codes. The experimental results shows that this model not only shorten the execution time of word segmentation program effectively, but also improve the grid resources utilization.