Discovery of Web pattern from web logs files using enhanced graph grammar approach

Searching useful information without fault from the Web becomes an increasingly difficult task, since the volume of Web data rapidly grows. With the growth rate, unexpected faults of Web service composition may occur in different levels at runtime. These faults are to be identified from Web Log files. The common causes of faults in Web services execution are rectified by fault diagnosis technique. So far, most existing approaches focus on the log content analysis but ignore the structural information and lead to poor performance. To improve the fault classification accuracy, fault classification analysis is carried out in this proposal. The Enhanced graph grammar algorithm is incorporated for identifying different types of fault categories in the form of graph structures.