1Integration of Collaborative Information Systems in Web 2.0 Ahmet E. Topcu1,2 , Ahmet Fatih Mustacoglu1,2, Geoffrey Fox1,2 , Aurel Cami3 1Community Grids Lab, Indiana University, Bloomington, IN, 47404, USA 2Department of Computer Science, Indiana University 3Department of Biomedical Informatics, University of Pittsburgh { atopcu, amustaco, gcf }@cs.indiana.edu, [email protected] Abstract We describe a new integration model that uses tools and services for supporting Web 2.0. This integration model defines a structure for a missing feature of Web 2.0. The model integrates a number of existing online tools having a common data model and aims to develop added-value community-building integrated environments. We discuss the overall design, architecture and the components of the integration model in the Web 2.0 domain. KEYWORDS: Integration, Web 2.0, information systems, collaboration, academic search. 1. Introduction The evaluation of the Web shows that people want to access information easily, store them in a personal way, and share them with the others. There are numerous tools and services built in recent years in different categories having Web 2.0 capability. Examples include Social Bookmarking Tools (YouTube, del.icio.us, Flickr), Blogs (blogger.com, Google Blog), Social Networking Tools (MySpace, LinkedIn) and other related tools. The users of these tools have the opportunity to use different tools and decide the best ones in their perspective. Users do not need to know about the version of the tools and services [1]. However, having many tools in similar areas is a problem. What if a user wants to use some other tools, how can the user move the data from the previous tool to the new tool? What if the user decides to use similar tools in the same environment and compares information at the same time? In other words, users should have a flexible environment to use multiple tools at the same time. A possible solution may be to define an architecture defining a model for integration to combine similar tools and use multiple services to user community to solve this problem. The web technologies such as RSS (Really Simple Syndication) [2], ATOM [3], AJAX (Asynchronous JavaScript and XML) [4], microformats [5], and REST (Representational State Transfer) [6] provide flexible Web-accessible data and services for Web 2.0 applications. However, although the current systems are for the most part good, they are independent of each other. Huge amount of data distributed over different tools and services exists in the Web. So, in this rapid development cycle one tool might have an advantage to the other tool and vice versa. One of the features of Web 2.0 is the focus on the people. The platform is motivated by questioning how people should interact with each other and easily share data in the Web. The resulting tools are easy to use, and allow people to put information and download them easily. However, there is no such a mechanism to combine them and have richer data or metadata integrated services. This model is created using native tools and wrappers around them without re-building the tools. Also local capabilities for example local search capabilities can be added and embedded in different client models, such as gadgets. So, the model has the capability to upload information to the tools and download information from them. The model should also provide sharing of logging of users. In this paper, we describe an integration model and its components for Web 2.0 using web-accessible data and services. This model is motivated by the above concerns to provide flexible mechanism to integrate similar Web 2.0 tools which have similar data model. The rest of this paper is organized as follows. Section 2 gives an overview of the existing online tools in the Web search domain. They are integrated in the Semantic Research Grid (SRG) prototype model. Section 3 describes the architecture and components of the integration model which is defined for the Web 2.0 platform. Section 4 provides an overview of SRG and its modules. Section 5 concludes this integration model. 2. Overview of the Web Search Tools In our integration model, which will be discussed in Section 3, we use some of the major open-access academic search tools. These tools are discussed next.22.1 CiteSeer CiteSeer has pioneered a number of techniques for the automated extraction of document metadata, including front-end metadata such as title, author names, author affiliations, abstract, and back-end metadata, such as acknowledgements, and citations to other papers. The algorithms used by CiteSeer are generally based on carefully crafted heuristics and/or machine learning techniques. Recently, it was estimated that CiteSeer covers about 24% of papers in Computer Science and it was pointed out that the use of automated methods for harvesting documents has led to a bias toward papers with 3 or more authors [7]. To deal with issues such as increasing query latency and degradation of system stability, as well as to improve the interoperability of the system, CiteSeer has recently announced the design of a new version of the system, called CiteSeerX [8]. 2.2 Google Scholar GS has been generally lauded for the open, fast and easy access it provides to vast collections of digital academic documents. There has also been significant criticism towards GS, especially from librarians. The major criticism has to do with: (i) scope (GS does not declare which publishers it currently covers; at the same time it is known it does not cover some major publishers, such as Elsevier, American Chemical Society, and Emerald [9, 10]); (ii) coverage (GS does not provide full coverage of the articles from the publishers that seem to be covered [9, 10]); (iii) accuracy (its metadata extraction algorithms are not very precise, leading to duplicate records, unreliable citation counts, etc. [10]). 2.3 Windows Live Academic Windows Live Academic (WLA) is one of the online