When most people think about “semantic” or “Semantic Web”-based software, they tend to think about applications that are quite explicit about their use of RDF, SKOS and OWL. While these types of applications are clearly becoming more popular, the vast majority of people have no clue why they should care about such things.

A new form of semantic application is emerging that benefits from these rich data sources, but isn’t showy about it. It translates the content into a powerful, easy-to-use consumer-oriented application. Many would argue that this is how it should be. Average users don’t care about HTML per se, they care about documents and web applications. They don’t care about relational databases, they care about software that allow them to view, query and manipulate the data. XML is not something that excites them, but robust, validated information exchange between systems is (or at least should be).

Smart phones (and by extension, tablets!) have become the most exciting new platforms for consumer-oriented applications. It isn’t surprising, then, that this is where we are starting to find innovation around semantic technologies. As ubiquitous devices that are always with us, they have the most information of how and where they are being used. Time, geolocation, personal identity and interests and application context are all rich fodder for integration with other data sources, public and private. While the overall number of examples of these systems is still small, they are being hailed as game-changing examples of what is to come. And, the number is growing. Here are a few examples.

Virtual Personal Assistant: Siri

At SemTech last year, Tom Gruber introduced Siri, a new form of virtual personal assistant. Sporting both a slick graphical user interface and a Nuance-powered voice recognition interface, Siri astonishes at how fluid it is to find things you are looking for. Siri is able to answer requests such as:

• “Find me a romantic Italian restaurant near here.” 
• “What’s going on around here this weekend?”
• “Find funny movies that are playing near my home this weekend.”
• “Remind me in ten minutes to go outside.”

The voice recognition works so well for simple (but non-trivial) commands, it is a little frustrating when it falls short of standard issue 24th Century Federation kit. Ultimately, it is the flexible backend data models and the service delegation infrastructure that maps data references to relevant services that makes it so exciting and useful.

Exploring Connections: EvriVerse

Evri has released a concept exploration application called EvriVerse. Concepts are related to each other based on current news activity and long-term relationships. Each concept is visually connected to related concepts in an easily navigable form.

Behind this simple user interface, EvriVerse uses the same API the Evri platform provides to Web developers. According to the company’s backstory, EvriVerse grew out of a 20% project but has become an important showcase for the kinds of things the company does well:

• analyze text of unstructured content (web pages, Twitter feeds, etc.) to identify entities
• extract entity networks of related concepts
• categorize documents
• receive content recommendations based on the context of the request

Hopefully, the company will continue to invest in the app to add more capabilities and more external data sources to link against. I have not heard any indication if and when the Android platform will be supported. If you know, please let me know.

Semantic Note-taking: Snaptic

Snaptic represents a new, richer vision of semantic note-taking with its 3Banana Notes application. The idea, as with the other apps mentioned here, is that the act of taking a note is done within a context. That context could be temporal, geospatial or within another application. This metadata can be captured about the notes to help organize them. Additionally, users are able to specify metadata within the content of a note that will assist in its organization (think Flickr or del.icio.us folksonomic tagging). All notes are private unless shared explicitly by users. Notes that are shared can be organized based on their related metadata.

One of the real insights is that capturing content from arbitrary applications becomes just another form of note-taking. Capturing a geolocation from the Compass application can be a note. Saving information from a recipe application can yield shopping lists as another note. Snaptic is partnering with developers to produce many such integrations as shown in their gallery. Developers can find out more information here.

Snaptic’s 3Banana Notes is currently available on both the iPhone and the Android platforms as well as the Web.

Summary

The three applications mentioned here are not brand-spanking new, at least not new this week or month. It was the context of the excitement about the iPad release that got me thinking about how that platform might affect the semantic application space. In the process, I realized that there were already good things afoot and worth discussing. The main observations under discussion here are not terribly profound, but they are important.

The first point is obvious, but often ignored. Information is produced within a context, and capturing that context helps empower users to reuse the information outside of the context in which it is produced. This is one of main promises of Semantic Web technologies and it is heartening to see it showing up in consumer-oriented software.

The second main point is that a common feature of these applications is a simple, accessible user interface. It is unclear exactly where semantic technology use begins and ends and the users are probably the better for it. Developers should not be afraid to sweep the details under the rug from time to time. Applications that benefit from complex and rich data sets might also benefit from not exposing these details to their users (or at least all of them). Simplicity and clarity in the user experience might be compelling enough when rich data models are used behind the scenes.