Recently launched in the Apple store for iPad tablets is a free version of Epsilon, a semantically-enriched mobile e-learning app from SemantiStar, the company that was previously involved in the development of the iPad news reader app StreamGlider. (See our coverage of that product here). Epsilon, as SemantiStar CEO Bill McDaniel puts it, provides an example of the company’s primary goal, which is to put semantic technology into traditional enterprise applications.
Building on research SemantiStar was involved in with DERI a few years back about what semantic pieces of technology would make e-learning better, the Epsilon mobile app is positioning itself to be the delivery vehicle for enterprises’ courseware, regardless of what their learning management system or repository may be. It has core support for Moodle format courses, but Epsilon can be attached to any LMS as long as there is an open API. The free, downloadable version is the entryway to the subscription-based enterprise version available directly from the company.
The app includes the ability to select courses directly from a variety of repositories for download and delivery; support online and offline learning; and facilitate study through built‐in quizzes and other features. Administrators can define a course map and users can build a personal catalogue of courses required. It includes a multi-level course semantic and keyword search tool and the Relevant Knowledge feature, where its semantic talents are really highlighted. The feature connects learners contextually with course material that it interprets and real-world knowledge and events.
“As you consume the course, on each page we extract keyword entities out of the page and present them in a tag cloud, from most relevant to least,” says McDaniel. The app goes out to the web to pick up news or Wikipedia articles relevant to those now disambiguated entities – and relevant in the context of the course, as well – in realtime. Popups and short summaries of the information are presented to help establish context for the user about the learning material. “So we try to understand what the course is about, then pick up keywords that are relevant because of the content, then present articles,” he says.
As an example, included in the sample free app is a course on the novel Dune. “Because we know the context is not dune as in a desert of beach [but Dune the novel], we find Dune as the significant, relevant keyword and look up articles about the book,” says McDaniel. “That happens on every page and in realtime. So, literally, every time you turn the page you could get a different article or different set of references, and you can drill and expand on those and see them in your browser.”
McDaniel says work the company did last year for an educational company working with a for-profit university to map job skills to job postings helped with developing Relevant Knowledge. “We had the same problem here, where it’s fairly sparse bits of text – only information from the page of content and the entities we extracted [to work with] – and you have to look and match [articles to them] in a relevant way,” he says.
Any concerns about corporate elearners picking up real-world stories the company might prefer that they avoid? About its president under investigation for, say, compliance violations for a user taking a course on the topic of compliance? No worries, McDaniel says: The administrator is able to control the list of RSS feeds they want searched and he or she can turn off access to Wikipedia. “What the feature does do is add a great deal of immediate context to what may be a rather dry course, which helps in retention, understanding, comprehension and awareness,” he says.
The enterprise version will include customizing the software to the buyer’s LMS, setting up a distribution mechanism, and support. The company also is working on a semantic-powered knowledge-sharing collaborative enterprise product, to be called Eridanus. McDaniel is hoping to have that available by year’s end.