FST Graduate Research Topic

Research Topic: A system to embed social interaction everywhere
Suggested By: Jon Dron
Last Update: May 16th, 2025

Overview   Most organizations run multiple websites for different purposes, some of which have social tools, others that don't. On the whole, there is little continuity of social presence from one to the next: they might, say, use a Discord site, a group email site, a public website, an accounting package, a learning management system, and so on, each of which is either asocial or only minimally connectable with the rest, through (for instance) social media or email sharing links. This project is concerned with building a social substrate that potentially pervades any and every website used by an organization, no matter how diverse, so that conversations and other interactions can occur anywhere and everywhere. Metaphorically, this would make an otherwise independent set of websites behave much like a city, where many disparate buildings and locations are bound together by the communities that form around and between them. No matter how diverse the sites in functionality or look and feel, visitors would feel that they were part of the same, broad community when visiting any one of them. Note that there are a few partial solutions to this problem already. Disqus, for example, provides a widget that can be embedded on most public pages to allow comments to be added, that are stored on a centralized site, which provides services such as authentication, authorization, comment moderation, and social networking. It is aimed almost exclusively at public-facing sites that need to engage an audience. it is mostly used as a blog commenting tool, allowing individuals to add an interactive conversation to a wide assortment of systems. It is not very suitable for intranet or partially closed sites, where there are existing mechanisms to authorize and authenticate at least some of the users. It is also a closed, proprietary system that requires people to have an account in order for it to be used effectively. Similar problems affect other vaguely similar tools, like those from social media sites. An organization such as Athabasca University cannot insist or expect that its members and visitors should have an account on a particular social media site, especially one that they do not control that has its own agenda and profit motives. It can, however, expect that its members will have an account on internal systems, and it can manage its own user database of external users. Problems to solve Among the many issues that need to be resolved are: How best to integrate such a system with a closed, or partially closed community site that already provides social interaction. For instance, how might it integrate with Brightspace discussion forums, SharePoint, or Athabasca Landing? How to build embeddable code that can run on any and every web page, no matter how it is built, without disrupting the functionality and without breaking the formatting of the page. Note: framesets are one possible solution, and might provide a baseline fallback when all other methods fail, but they are not a good solution! How to handle access rights and permissions, including for anonymous visitors to public sites, in a manner that is controllable by a page owner or owners, that is secure, that is low-threshold, that does not reveal any information to anyone not authorized to see it. Note the need for control: page owners should be able (for every single page) to make discussions as open or as closed as they wish, from totally public to just one or two named people. On some pages there may be a need for a mix of public and closed discussions. Note also that access rights may be set locally (e.g. only some staff can see pages relating to leave requests or finance, only course members can see a Moodle course, etc) and the system needs to accommodate (and ideally reuse) this. How to build a flexible but easy to use interface both for those involved in discussions, and for the owner of a page. How to leverage and/or aggregate discussions - for example, to collect/find pages on which an individual has commented, to recommend other pages, to identify cross-overs, to highlight popular discussions, to uncover pages that have fallen into disuse, etc. How to handle ownership and notifications: how to deal with changes in page ownership, multiple page owners, orphan pages, etc. These are just a few of the most obvious issues: students are encouraged to explore this problem space, and to seek creative solutions. There are also opportunities for doing part of the research that could help drive this forward, for instance into interaction design, technical standards, available platforms, and so on.   Example partial solution and possible starting point A social bookmarking tool, such as that used for the Landing (an Elgg-based system), provides the ability to share and comment on a site, and for discussions to form around it, as well as other tools for organizing bookmarks, rating them, and forming social networks or groups containing people that may use them: very importantly, a system like Elgg can handle discretionary access control, from fully public to fully private, very easily, and can integrate multiple authentication sources quite transparently, including single-sign-on/authorization standards like CAS, Shibboleth, OpenID or Active Directory. It can also handle issues like moderation, spam prevention, and notifications. In brief, though not embedded directly on the bookmarked pages, it provides much of the required functionality out of the box. Normally, such a site would be a stand-alone centralized repository of shared links, accessed by users of the site and visitors to it. However, it doesn't need to be that way. A framework like Elgg can also present itself via an API so, for any given page, the comment and discussions could be drawn from the site and embedded in that page, as though it were part of it. This would probably work very well for simple HTML pages on a public site (much like Disqus), and would provide plenty of tools for integration and aggregation on the central bookmark repository, not to mention social networking, user management, etc. However a range of thorny issues (see problems to solve above) would remain, from technical embedding issues to integration with existing site communities, to social/organization problems.