What will you learn

Prerequisites

Operational Matters

The term Semantic Web is credited to Tim Berners-Lee.

"Semantic Web -- a web of data that can be processed directly or indirectly by machines" -- Tim Berners-Lee in Weaving the Web.

However, research in "Semantics," particularly Semantics of Information as is relevant to the Web as the global information system, predates the Web. Semantics has been studies in many fields, and from information systems perspectives, we will use Semantics to mean "meaning and use of data."  With this mind, we view Semantic Web as the concept that Web-accessible content (information) can be organized and utilized semantically, rather than through syntactical and structural methods. 

So this course is about investigating the next generation of the Web whose key distinguishing characteristics will be the support for and use of semantics in new, more effective, more intelligent, ways of managing information and supporting applications.

Top

What will you learn?

Here are some of the things you can expect to learn (some superficially, some in depth):

• What is Semantics? Syntax, Structure and Semantics
• Understanding content: Metadata, metadata standards, XML+metadata specification, RDF and metadata processing
• Semantic underpinning: Ontology, Domain Modeling, Logic, Inferencing, Context
• Classification  and semantic metadata extraction techniques: statistical, statistical learning/AI, lexical and natural language, knowledge based
• Specifications: why is XML(S) not adequate? why is RDF(S) not adequate? what does DAML+OIL offer?
• Semantic Applications - demonstrating power of semantic technology for search, personalization, contextual directory and custom/enterprise applications; next generation semantic content management
• Research Landscape: review of some of the active projects (e.g., SHOE, OntoBroker, InfoQuilt) and initatives (OntoWeb, DAML)
• Commercial Landscape: For example, technologies and products from Voquette/Taalee, Ontoprise, Cyccorp, Applied Semantics, business models today and in future
• Research Questions: Contributions of IR, AI, Logic, NLP, DB and IS to Semantic Web, Ontology integration versus interoperation, Broadening the current vision of Semantic Web (beyond machine understandable data) to include modeling of human information and decision making needs

Top

Prerequisite (who may take this course, who should not)

This is an advance course, ideally suited for PhD students, MS students in their second year, MS students with work experience or MS/PhD students wishing to pursue research in focus area of the LSDIS lab. The student must have understanding of and expertise in the following three areas:

• database, information systems or AI principles (such as a course in database management, global information systems, AI, or work experience involving creation of significant database schema and data intensive application (it is possible for the student to focus on either database/information systems or AI aspects/technologies of Semantic Web),
• basic or advanced exposure to the Web, including HTML (must have created a home page), XML (self-reading may be ok for a good student), and
• very good programming skills, minimally in Java and any scripting language (preferably a course in Software Engineering or Compilers; just a course in data structures will not be good enough)

I particularly welcome students with lots of curiosity as well as research minded students. This course is not suitable for students who prefer highly-structured, text book based education. Semantic Web is a fast evolving field -- each day I will find new research directions, semantic applications, W3C standardization efforts, technological advances and products. I want to be able to share these with the students.

Top

How will Course be conducted?

There will be three components of the course:

• instructor's lectures
• assigned readings, followed by classroom discussions or student presentations
• small and medium size research assignments
• a group project

It will be necessary for the student to attend ALL classes, except for legitimate health  reason (typically requiring doctor's note or a copy of prescription). Expect to spare time to read one chapter or paper per class at the minimum. Active classroom participation will be an important barometer for your grade. Grading components will include (a) classroom presentation, (b) class room participation, (c) project, and (d) possible a mid-term.

Educational Philosophy: I am interested in seeing you "learn how to learn" rather than just learn information already laid out by others.

Top