NCeSS - Proramme

Events

NCeSS Conference

Content Repositories and Social Networking: Can There Be Synergies?

Dolby, Paul; Fish, Adrian

This paper details the novel application of Web 2.0 concepts to current services offered to Social Scientists by the ReDReSS project, carried out by the Centre for e-Science at Lancaster University. We detail plans to introduce Social Bookmarking and Social Networking concepts to the repository software developed by the project. This will result in the improved discovery of e-Science concepts and training to Social Scientists and allow for much improved linking of resources in the repository.

We details plans that use Social Networking concepts and Open Social Networking standards which will promote collaboration between researchers by using information gathered on user's use of the repository and information about the user. This will spark collaborations that would not normally be possible in the academic repository context.

In developing this software a number of issues are raised, in particular confidentiality when sharing user's information and the quality of bookmarked resources. These issues will be discussed and our solutions will be presented.

This paper will demonstrate how adopting these Web 2.0 concepts in a repository can benefit other academic repositories. By adopting Open Standards, this adoption can be simplified.

The Resource Discovery for Researchers in e-Social Science (ReDReSS) project[1] is a JISC/ESRC funded project aimed at stimulating the use of e-Science in Social Science research. The project, which has been running since October 2003, employs a small team of developers to produce and harvest e-Learning content to support this aim. One of the many resources offered on the ReDReSS project's website is the Learning Space Catalogue[2] (LSC); this is the projects main vehicle for dissemination of its training and awareness raising material.

The LSC is a metadata repository that holds the records of hundreds of e-Social Science learning resources. Many of these resources have been developed by the ReDReSS project or harvested by the ReDReSS team from the internet, on account of their quality and relevance.

The LSC provides a number of tools to allow Social Scientists to easily discover resources that may stimulate their uptake of e-Science and to provide them with further training. One such tool is the Bookmark tool. This tool allowed users to bookmark resources that they find of interest. These bookmarks can be saved or printed once they have finished browsing the LSC. This paper describes the work being carried out to enhance this bookmarking tool using Web 2.0 concepts, in particular Social Bookmarking technologies. This new tool will promote the discovery of e-Social Science resources. It will also promote collaboration between researchers.

Social bookmarking provides a method for internet users to store links to web content that is primarily of interest to them, but may well be of interest to others researching similar fields; unlike tools provided within browsers to store bookmarks which remain personal to the user, social bookmarking web sites allow users to share their bookmarks with other users.

Bookmarking tools built into browsers allow users to store their bookmarks in folders. Social Bookmarking web sites such as such as Delicious[1] allow users to tag their bookmarks with textual labels which they think are descriptive of the resource being bookmarked. Tagging is a process whereby users associate keywords with the particular bookmark. They can then build up a set of tags which are words or phrases that have a particular meaning to them. All bookmarks with the same tags can be grouped together, allowing for easy and quick retrieval of bookmarks which are of interest to the user. Since the user is tagging with words that are familiar to them, rather than using a rigid set of predefined tags (taxonomy), information retrieval becomes much easier for the user.

Once tags have been assigned to a bookmark a user can make these tags and the bookmark available to other users. The other users of the social bookmarking site can then find bookmarks using the tags created. They can then provide tags of their own for these bookmarks. Through this tagging process a folksonomy gradually develops that eventually produces a set of common terms that can be linked to the bookmarks. This folksonomy allows web resource to be easily discovered by users. Unlike traditional web search engines, this folksonomy is created by humans who are more likely to understand the content of the web resource. This has advantages over machine classification.

Social bookmarking web sites such as Diigo[4] go a step further and allows for fine grained tagging. Diigo is a social annotation web site and allows users to tag/annotate particular text on a web site.

In 2004, with the introduction of Connotea[5] from the Nature Publishing Group and also Citeulike[6] social bookmarking was made available specifically for researchers. Both are online reference management tools allowing researchers to manage their references using tagging. Connotea also allows for the automatic extraction of metadata, such as creators and publishers, from the online resources. Both of these bookmarking sites, however, are typically aimed at bookmarking papers.

The first section of this paper we will provide a brief but more detailed overview of social networking to date, expanding on a review carried out by the Nature Publishing Group in 2005[7]. We will then describe the application of some of these concepts to the LSC.

Allowing users to bookmark and tag resources in the LSC will enable the generation of links between resources that will offer other users sets of resources that will aid their understanding of e-Science subjects. These links will be much more accurate than a machine automated semantic method for linking resources.

In the majority of this paper we will be concentrating on the implementation of one particular aspect of social bookmarking to the ReDReSS LSC, the social networking aspect.

An important aspect of e-Science is collaboration. Through gathering information on the users of the LSC and their bookmarking habits we can enable users to build up a social network linking users with similar interests together, sparking collaborations that may not have occurred otherwise.

A number of issues occur, including confidentiality, when sharing information on users. In this paper we will address these issues.

There are also a number of issues that occur when relying on users to tag and add bookmarks. For example there are issues concerning the quality of the material being bookmarked. This does not present a problem if we restrict users to bookmarking only the content contained in the LSC. The LSC content has already been assessed for its quality by the ReDReSS team. However allowing users to add their own bookmarks or even bookmark their own content, will present a quality problem. This has led, for example, to the creators of the Wikipedia[8], a social encyclopaedia web site to employ experts to act as editors to the content[9]. In this paper we will discuss present possible solutions to this problem.

Throughout the development of the LSC software we will use open source code where applicable, to save on development. We will also adhere to relevant standards. Standard become important when sharing information with other social networks and repositories.

Our repository already uses the Dublin Core[10] standard for its metadata, which is common amongst repositories. There are plans to implement OAI-PMH[11] services to allow the sharing of metadata between other repositories. It is not common, in the academic world at least, for repositories to share information about users. In November 2007, in the commercial world, Amazon and the social networking web site Facebook started sharing (controversially) some user information. This was information such as details on the books that they have bought. It is not however clear if common standards we used in this process.

In this paper we will discuss appropriate Open Social Networking standards and our implementation. Such standards include FOAF[12] (Friend Of A Friend) and XFN[13] (XHTML Friends Network).

Using common standards to express information gathered about users and their bookmarking habits, we will demonstrate in this paper how we can develop links between people using our repository, other repositories and social networking web sites. This has a number of advantages. It will promote collaboration between Social Scientists and it will expose content in the LSC to new audiences and provide links to new content for the repository.

The benefits of the techniques described in this paper to other academic repositories will be discussed in the final section of this paper.

[1]http://redress.lancs.ac.uk/get.php?41

[2]http://redress.lancs.ac.uk/Learning_Space/

[3]http://del.icio.us/

[4]http://www.diigo.com

[5]http://www.connotea.org

[6]http://www.citeulike.org

[7]http://www.dlib.org/dlib/april05/hammond/04hammond.html

[8]http://www.wikipedia.org

[9]http://www.theregister.co.uk/2005/12/19/sanger_onlinepedia_with_experts/

[10]http://dublincore.org/

[11]http://www.openarchives.org/OAI/openarchivesprotocol.html