Research focus: Semantic Email

PROBLEM

Effective email management is simple with a reasonably small mail set. In the world of bulk email, this is quite the contrary. Email tools degrade in utility as the number of messages grows. Inversely, searching, sorting, filtering, and blocking of SPAM become steadily more important as the volume of emails increases.

To manage a dynamically increasing volume of emails, mail client users need a content based indexing system. I will use the budding Semantic Web as the backbone for scalable email tools that utilize the semantic content present in messages. I will study the effects of applying the Semantic Web to large email databases (be it personal mail or an archive), identifying any benefits as well as describing the feasibility of such a merger.

PROPOSED SOLUTION

Searching, sorting, and filtering within a dynamically increasing volume of messages requires tools that work on semantic data. The difficulties lie with embedding and utilizing this semantic markup.

Embedding metadata:

• Current email protocols do not support embedded RDF, the language of the Semantic Web. My task will be to design a protocol which is MIME compatible while containing the necessary RDF syntax.

• Physically embedding the RDF is also a problem. Without increasing message generation time significantly, I must provide the sender with tools to create a second set of information, namely the metadata. I will create a markup tool that streamlines and simplifies this process.

Utilizing metadata:

• I will develop email tools that manipulate the semantic content present within a given email database. Searching, sorting, and filtering are the primary functions that stand to benefit from semantic connectivity, as well as a range of lesser utilities.

• I will create an HCI to facilitate user testing. Productivity increases will be measured through experimental testing of different interfaces built on semantically keyed tools.