Workflow Application Of Clustering Tree Edit Distance

The workflow is a reflection of a computerized model of the business processes. It plays an important role in areas such as office automation system, e-government system. It also can help enterprises to realize the business process modeling, business process optimization, etc.With the extensive application of workflow, workflow model library appeared, Workflow in it was updated along with the changes in the real demand and more and more workflow was added to the library.Therefore, So a systematic way to analysis and to improve workflow becomes very important. In recent years, many scholars begin to pay close attention to the workflow finding and workflow clustering.The workflow Clustering is a process which makes the physical or abstract collection grouped into multiple clusters composed of similar workflow process. Introducing workflow clustering, we can search and analyse the workflow models in the library quickly and accurately. The structure and content which conveyed by the workflow between each other is different, so it is necessary to find a new workflow clustering theory and method suited the character of workflow to meet the needs of the business.We give a brief introduction of the management of workflow models and workflow clustering. First,according to the character of the workflow we propose a quantitative measure to calculate the distance between different structured processes. In this way, we transform each workflow into a process structure tree, and then calculate the workflow distance based on the tree edit distance and word semantic distance of two structure trees. This method is feasible. Then we adopt hierarchical clustering method to cluster the workflow models in the library.Finally, the article completed a system supported workflow clustering model.Confirmed by the experiment, the workflow clustering method can accurately get the clustering collections meeting the needs of users, and widely used in the areas such as process reengineering.