An approach and tool support for assisting users to fill-in web forms with personal information
Résumé
Web forms are massively used as a very effective way for user interaction with information systems. Notwithstanding, filling in forms with personal data can be tedious and repetitive. Due to legal and technical constraints, full interoperability of information systems is not a straightforward solution. So that several client-side techniques have been developed in the last years to automate the task of filling in forms; for example, auto-filling and auto-complete are very well-known techniques that employ contextual information to fill in automatically Web forms. However, the accuracy of these techniques is limited by the contextual information available on the Web browser. Some information systems can record users' personal information on the server side and use them to provide pre-filled forms to returning users. The problem with such as an approach is that users must keep updated records of personal information in remote servers; legal and technical issues prevent from sharing personal data among different applications, thus users must maintain multiple accounts. Interestingly enough, the analysis of data requested in forms reveal a pattern in the set of pieces of personal information that are often required (e.g. names, affiliations, billing address, home address, bank account, etc). In this paper we propose a new approach for automating filling in form that relies on these patterns of personal information. Our ultimate goal is to provide means for supporting the exchange of data between user's Personal Information Management Systems (PIMS) and Web forms. The approach is supported by a tool called PIAFF (which stands for Personal Information Assistant for Filling Forms) and illustrated by a case study concerning forms used for student applications.