GUIDE: an interactive and incremental approach for crawling Web applications

摘要

The Internet, having a sea of Web applications, is one of the largest data stores for big data analysis. To explore and retrieve the states (pages) from Web applications, Web crawlers have been extensively used. Most crawlers allow the users to define a few crawling directives so as to increase the coverage of states that the crawler can explore. A directive can, for example, assign an input value to a specified input field so that the application is instructed to perform a specific action and visit some special states. Note that, a crawler is supposedly capable of exploring an unknown application. But, given an unknown application, how could the user possibly prepare the required directives in advance? This paper proposes an interactive crawling approach and a crawler called GUIDE to overcome this issue. Instead of passively receiving directives from the user, GUIDE actively asks the user for directives when Web pages containing input fields are found. In addition, GUIDE offers a hierarchical directive structure, allowing the user to define multiple values for the same input field. A case study with three Web applications indicated that (1) interactive directives were very useful for increasing the code coverage of the application being explored-up to 10.3-50.5% of code coverage improvement can be achieved, and (2) using GUIDE is more efficient than using a traditional crawler-given the same amount of time, up to 11% of code coverage improvement can be achieved.