At the NSDL 2009 Annual meeting, Folksemantic team members participated in a Panel on recommendation systems. In that panel, we addressed the following questions:
Personalizing learning and teaching support has the potential to make them more efficient, effective, and enjoyable. Systems support personalization by adapting functionality to individuals, by allowing users to customize the system, and by supporting human interaction inside of the system.
In creating OER Recommender and OCW Finder our goal was to provide a means for educators and students to easily find and access NSDL and OpenCourseWare open educational resources (OERs) that meet their individual needs. The OER Recommender widget added the ability for users to discover "like" resources while exploring pages of interest. By recruiting partners to provide metadata for pertinent and vetted educational resources, we continue to expand the range of content that can be found using our services.
Adopting the perspective that users want a more personalized experience than what search engines offer, we have expanded our search-and-retrieval system to incorporate social aspects. Under the umbrella of "folksemantic.com", we have integrated OER Recommender and OCW Finder and added functionality for users to connect to other educators and learners and soon to receive personalized recommendations.
Web sites can easily include our "related resources" widgets in their site to help their users find relevant resources whether they be on their site or others. Please see (http://www.folksemantic.com/widgets). We have implemented the ability for people to register on the site, and add the feeds that they produce to the system such as their blogs and bookmarks for display on their profile. This allows other users to learn about them and decide whether or not to "follow" them in the system. Users can easily invite their colleagues to participate in the system by importing their contact lists. When users login, they are presented with a Facebook-like dashboard that includes an activity feed showing the activities of the users that they follow. The dashboard will also be the place where the system will display personal recommendations of resources and people that may be of interest. Personal recommendations will be based on attention metadata including searches, click data, time on page, registered feeds including bookmarks and feeds, comments, and shares.
Our primary audiences are learners and teachers who access OERs in the NSDL and in OpenCourseWares; primarily middle schools, high schools, and post secondary schools. The audience of our recommender widgets are the OER repositories themselves.
We have not attempted to measure learning. Our index includes over 110,000 OERs including over 3,500 OpenCourseWares and OERs in 8 languages. Over 1400 users have registered in the past 5 months. We recently added functionality to allow users to register feeds they produce such as their bookmarks and blogs, and to share and comment on OERs in the system. We have conducted research on algorithms for personalized recommendations.
Folksemantic is a step in the direction of providing Open Distributed Learning Management Services that support teaching and learning with OERs. Our widgets and services approach "goes to where the resources and people already are" rather than attempting to bring resources into a learning management system and get people to come to a site. NSDL repositories can easily adopt our recommender, discussion and sharing services to add value to their sites. As we add personal recommendations, these will be pushed to the user via email, twitter, Facebook and other channels. We envision additional light-weight services including assessment, learning management, adaptation, and sequencing tools that can be adopted and integrated into existing environments with little effort. The resulting influence on teaching and learning will be that learning and teaching will be more readily integrated into normal web activities. In addition, online learning will be more interactive and collaborative.