G-Form: A New Approach for Visual Interpretation of Deep Web Form as Galaxy of Concepts Radhouane Boughammoura, Lobna Hlaoua and Mohamed Nazih Omri Faculty of Sciences of Monastir, University of Monastir Research Unit MARS, Monastir, Tunisia E-mails: Radhouane.Boughammoura@gmail.com, Lobna1511@yahoo.fr, MohamedNazih.Omri@fsm.rnu.tn Abstract—Deep Web is growing rapidly with multitude of devices and rendering capabilities. Despite the richness of Deep Web forms, their rendering methodology is very poor in terms of capacity of expression. Hence, the user has no indication about the richness of the query and the query capability when he interprets this interface. In this paper, we propose a new rendering approach of Deep Web forms which is easy to interpret by the user and reflects the exact meaning of the query. We have evaluated our algorithm on standard dataset and compared it to a well known state of the art algorithm. Our approach showed good performance with respect to standard measures. Keywords- Web Applications; Deep Web; Information Retrieval; Query Interface; Query Interpretation; Galaxy of Concepts; Pertinence of a Concept; Human Computer Interface; Visualization. I. INTRODUCTION Deep Web is the part of the Web which is not reachable via hyperlinks [2][4][7][12][14][15]. It is hidden behind Web forms which give access to Deep Web databases. Information on databases is a really big treasure. More than 90% of the information from the Web comes from Deep Web. In addition, this information is very rich in terms of quality of service offered to internet users [12][13][14]. We aim by our job to reveal the Deep Web to novice internet users via new, simple, and easy-to-use Web forms. A Web form is an information retrieval interface which give access to Deep Web data. It is a graphical representation of the query using a set of fields. Users form their query by visual interpretation of the meaning of the query interface. The design method of the Web form [1][3][5][6][10][11] is very important since it is the only source of inspiration for novice users in order to understand the meaning of the query. A bad interpretation leads to an incorrect query and hence restricts access to Deep Web services. In this paper, we focus on the design aspect of Web forms in order to offer to novice users easy- to-use forms. In the Web form presented in Fig. 1.(a), we notice presence of white fields crossing the Web form horizontally. These fields are very important; they indicate the presence of semantic entities (or semantic concepts).A novice user may not pay attention to these fields. We present below the relevance of the white fields. Let us consider an example in which the user is not interested in non-stop flights but is searching for a flight with one stop to reach the destination. Suppose the user is searching for all flights having as destination ‘’Tunis’’ offered by an airline company with one stop city. The user may give by mistake the destination city and the number of passengers and leave the departure city empty. This request will be wrong unless the user knows all stop cities for destination ‘’Tunis’’. This is not evident and will be a burden for a novice user as he must formulate as many queries as there are stop cities. In this paper, we present a new design methodology which simplifies the design of the Deep Web form. While our methodology preserves the query capability of the Web form, it removes the complexity of the query with an easy-to-use form containing all necessary and pertinent fields. The resulting form becomes very simple and, more importantly, semantically very rich. The rest of the paper is organized as follows. Section II presents a brief review of related works. Section III explains the motivation of our new approach G-form. In Section IV, we detail the principle of the G-Form and our experiments will be presented in Section V. Section VI concludes the paper. II. RELATED WORK According to the literature, Z-form [9] is considered as the most used form in the Deep Web. Z-Form is a flat query where all fields are listed at the same level of granularity. The name Z-Form comes from the fact that the user reads Z-Form like reading lines in a paragraph: he begins by the first line, then the second, etc. This reading strategy resembles to the letter Z (see Figure 1.b). 32 Copyright (c) IARIA, 2015. ISBN: 978-1-61208-437-4 INTELLI 2015 : he Fourth International Conference on Intelligent Systems and Applications