Abstract: For model verification, an approach to extract FSM models of Web applications is proposed. The extraction process is divided into three phases: link extraction, model construction and document generation. By using a regular expression and location technology for tags contained in a tag library built in the lab, information related to links is first extracted from the source codes of Web applications. It is then structured and stored in databases. In model construction, information obtained in the first phase is reconstructed with a mapping and aggregation technology after analyzing elements for model construction such as pages and links within Web applications. Finally, FSM models are obtained to verify the Web applications.

Key words: link extraction, Web application models, model extraction, finite state machines(FSM)