A Novel Approach to Aggregating Various Word Similarity Measures

In the recent years, the number of web services has risen up swiftly. Numerous works have been done on how to compose services. In the process of composing services, service matching plays an indispensable role. The importance of searching the most suitable service among composition can not be overemphasized. In order to find the best match for a service, the essential information in the service document should be extracted impact. Then the data should be put in a structure that describes the service perfectly. Then the difference between two structures from two service documents should be quantized. By using these values, a proper web service match discovery could offer a search result to the user. To measure the difference, word semantic must be considered. Many works focus on the performance of various measures for different tasks. Rarely do the researchers study on aggregating different measures. In this thesis, we propose a framework which aggregates different semantic measure for data extracted from WSDL. This framework is designed to identify the features in the service document, and use several measures for precisely interpret the difference between both semantic and structure information of two services.