1) Broad Overview - what is this all about - the W3C semantic web layer cake; what did the Scientific American article really talk about? Information architecture, past and future. This lecture is about setting context. What are the foundation technologies and who are the contributing communities? Convergence of web technologies, AI machine decidable logics, markup language and manipulation efforts, all over the last decade. Data models, data exchange, abstracting across data, machine 'understanding' of web. Where do these standards come from, and what are they driving toward?
Invited Speaker - TBD -W3C
2) Markup - annotating information, adding info to info; XML for markup, why do this, where does it fit in to the semantic web ideas? What is it being used for? What are XML's strengths, limitations; is it all we need for information reuse? E.g., "Talking about (real?) things: XML: A registration document contains a single license number field. RDF: A car has a unique license number." Semantic data is potentially much more reusable, and is about things in the 'real' world, not just things in documents or databases. What does that mean? Projects - talking about projects for class and requirements for students.
3) RDF - what is it and why is it needed? What is semantic data? Describing relationships in data; markup for metadata; semantic markup for web pages - inline, or in referenced databases - SI site example; power of shared standard - network effects of rdf described data that semweb is built on; will it succeed? How is it being developed? The community of standards development at W3C.
Invited Speaker - Eric Miller, head of W3C Semantic Web Activities Area
4) Examples of the use of RDF. Spending more time with fundamentals of semantic markup, and then starting to put what we have learned into practice. Some quick examples to show the value of this approach. Why do I want machines reading my pages? Friend of a Friend, creating machine readable home pages. Semantic blogging, adding semantic info to items shared over the blog channels, allowing navigation and search along semantic rather than simply chronological or serendipitous connections. Personal Information management using RDF - haystack project: a "universal Information client". The School of Information semantic markup project, intro to using TAP at SI.
Demos and discussions with developer(s).
Invited Speakers: tbd (from MIT haystack team)
5) Ontologies - what are they and how are they alike/different from things we are familiar with, e.g., taxonomies, relational databases and relational db entity diagrams, universal modeling language (UML); allowing enriched relationships descriptions; how to make them with Protégé; who is using them (google, yahoo, etc).
6) Tools for Ontology construction - writing your own and/or using an editor (like early days of HTML). Notation3 - an easy way to write down relationships; "RDF for sketching"; tutorial on N3. Protégé, a GUI for model construction; tutorial on Protege. Using existing schemas, models, ontologies; building on others' work and seeing how it contributes to interoperability. Examples, lots of examples.
7) TAP project at Stanford in depth. Demo and application for SI pages; Science of Collaboratories ontologies - use in research to organize complex info spaces - useful in analysis? investigating use of knowledge representation methods to formalize research descriptions. Example from SOC team.
Invited Speaker - TBD -on TAP Project (and/or on openCYC, the open source version of the Cyc technology, probably the world's largest and most complete general knowledge base and commonsense reasoning engine; may save that part for week 9.)
8) AI history and development, convergence with semweb ideas. This is the overview of the development of formal logics and their application to reasoning systems. Description logics, developments and applications in semweb ideas, problems. The view from the AI side of the house. DAML, OIL and now the OWL ontology languages; moving up the semantic web layer cake.
Invited Speaker - Ian Horrocks; OIL, DAML-OIL, OWL author/contributor
9) Reasoning engines - Cwm, the W3C reasoning engine, a general-purpose semantic web data processing tool (www.w3.org/2000/10/swap/doc/cwm.html); playing with the RDF models we have built, using some N3 forms of statements, and investigating the ideas of reasoning over them; working with research software and available tools; example of Connaly travel reservation system built with RDF and cwm. Analyzing SOC ontologies with a reasoning engine, perhaps. First steps in determining proofs on the web: systems should be able to explain their actions, sources, and beliefs. Time spent exercising tools and gaining experience with concepts behind them.
10) Semantic Web-enabled agents - The web as a global database and distributed computer, vs. the web as an information store. What are agents in this environment, how are they realized through web services, what do ontologies have to do with all this, and what's the difference between getting a reference to cattle populations and asking the web a question: "how many cows are there in Texas?" Realizing an example from the original Scientific American semantic web article: having software agents and web-based services organize a series of hospital appointments, map out the transportation to them for you and your parent (the patient), figure out how to share the transportation burden with your sister, what fits in whose calendar, and suggest where to purchase prescription medicine on the way home.
Invited Speaker: TBD, UMaryland
11) Applications in Education - Edutella; IMS/RDF/P2P and the construction of a semantically informed, distributed resource base for educational objects. The debates within the IMS (www.imsproject.org) community on XML versus RDF representations of semantic information.
Invited Speaker - TBD
13) Tracking an emerging discipline - Working across the disciplines of the web, the grid (the high performance international science network), and AI-based disciplines for knowledge representation/reasoning systems.
14) Project Presentations - we discuss projects students have engaged in that could include constructing concept maps, RDF models, RDF instance databases, or ontologies, and efforts at using the available tools for construction, and reasoning engines to investigate them.
Course books: (working list)