Knowledge representation and reasoning is the study concerning approaches and methods for symbolically representing knowledge and algorithms and techniques to manipulate such a symbolic representation in order to infer new knowledge from existingone. This field has received a lot of attention in recent years due to the emergence of Semantic Web, which has become a fertile ground for research and application of knowledge representation and reasoning. The central idea behind Semantic Web is to enhance data on the World Wide Web by so-called metadata, which describes the meaning (semantics) of the data. This enhancement is made possible by formal knowledge representation languages, which makes the data processable and understandable by machines.
In this course we cover in depth such knowledge representation languages for expressing metadata, called ontology languages. We will in particular cover the Resource Description Framework (RDF) and the Web Ontology Language (OWL), both of which are recommended standards by the World Wide Web Consortium (W3C).
Dr. Adila A. Krisnadhi (email: krisnadhi.2 [AT] wright.edu).
Office hours: Tuesdays 11am-12pm (please email first!) or by appointment.
Office is at 465 Joshi.
Tuesdays, Thursdays 12:30 pm - 1:50 pm in Russ 346
- Primary textbook: Pascal Hitzler, Markus Krötzsch, Sebastian Rudolph, Foundations of Semantic Web Technologies. Chapman & Hall/CRC, 2009. ISBN: 9781420090505.
- Ronald J. Brachman, Hector J. Levesque, Knowledge Representation and Reasoning, Morgan Kaufmann/Elsevier, 2004. ISBN:1-55860-932-6
- In particular, chapter 1, 2, 3, 4, 9, and 16.
- Documents mentioned in the Recommended Reading section of W3C's RDF home page.
- Documents mentioned in the Recommended Reading section of W3C's OWL home page.
- SPARQL standards (you can start from the Overview page).
- Other materials indicated in class as we go along.
There is no hard prerequisite for this class. Any knowledge about XML and about predicate logic will be helpful, but is not required. Necessary preliminaries will be covered in class.
Course outline (tentative, incomplete)
- Week 1 (8/30, 9/1): Overview, RDF
- Week 2 (9/6, 9/8): RDFS, RDF(S) Semantics
- Week 3 (9/13, 9/15): RDF(S) Semantics, RDF(S) Semantics
- Tutorial #1 Exercises [due date: 9/22].
- Week 4 (9/20, 9/22): SPARQL, Tutorial session #1
- Slides: 05 - SPARQL [extended and updated on 9/28]
- Week 5 (9/27, 9/29): SPARQL, SPARQL
- Week 6 (10/4, 10/6): SPARQL, OWL Syntax
- Week 7 (10/11, 10/13): Tutorial Session #2
- Week 8 (10/18, 10/20): OWL Syntax and Semantics
- Week 9 (10/25, 10/27): OWL Syntax and Semantics
- Week 10 (11/1, 11/3): OWL Syntax and Semantics, Tableau reasoning
- Slides 08 - Tableau algorithms for DLs [extended and rearranged on 11/10]
- Slides 08 - Tableau algorithms for DLs (with reduced animation, for printing) [extended and rearranged on 11/10]
- Week 11 (11/8, 11/10): Tableau reasoning,
- Week 12 (11/15, 11/17): Ontology engineering, Ontology design patterns, Linked data publishing with ODPs
- Readings for ontology engineering and design patterns: http://dase.cs.wright.edu/content/pattern-driven-linked-data-publishing-primer
- Week 13 (11/22): Linked data publishing with ODPs.
- Week 14 (11/29, 12/1): Linked data publishing with ODPs
- Week 15 (12/6, 12/8): Group project presentation
Homework 15%, Project 50%, Final exam 35%.
- Homework: Students will submit written solutions of the homework exercises to the instructor and take turns in presenting (their) solutions in class. Grading will be done based on the written solutions as well as evaluation on the participation in the tutorial sessions. (Grading dimensions: correctness, preparedness, understanding of the material, clarity of presentation, frequency of participation).
- Project: See below.
- Final exam: The final exam will be oral in the form of a 20-30 minutes interview with the examiner asking questions and the student answering. (Grading dimension: mastery of the course content). Schedule will be determined later.
- Group: 4-5 students.
- General task: develop an ontology to model a notion in a particular domain, populate with some RDF data.
- End products:
- the ontology (OWL file)
- an RDF document containing sample data that populates the ontology
- a project report detailing the ontology and the modeling experience while developing it
- presentation/demonstration of the project results
- Grading dimension: adequacy of modeling (based on the ontology, sample data, and sample queries), clarity and richness of the report.
- Topic choices (you may suggest other topics):
- Sport topics:
- American Football
- Board/Card Game topics:
- Chinese checkers
- etc. (Chess is excluded since I already co-authored a paper on this).
- Other topics?
- Sport topics:
- Milestone #1:
- Decide your group and topics
- For the topic you decide, list the possible real data sources for your topic. For example, websites the provide detailed information and data about your chosen topic. These data sources need not be already in RDF.
- Report to me your topics and group by October 18, 2016
- Milestone #2:
- Read Chapter 1 of the ODP handbook at http://dase.cs.wright.edu/content/pattern-driven-linked-data-publishing-primer
- List several competency questions pertaining to your topic. These questions should cover the use case of your ontology. See examples in Section 1.2 of the chapter. Take into account also the data sources you have identified previously.
- Based on the competency questions, list the key notions that must be modeled in your ontology.
- Describe in natural language the relationships between those key notions.
- Visualize the key notions and relationships between them as class diagrams. Consult figure 1.5, 1.6, 1.7, 1.8, and 1.9 in the chapter.
- Present your diagrams to the class on November 8, 2016.
- Final Milestone:
- Read Chapter 10 of the ODP Handbook at http://dase.cs.wright.edu/content/pattern-driven-linked-data-publishing-primer
- Produce an OWL file of your ontology that contains all axioms describing the key notions and the relationships between them.
- Produce an RDF dump containing example triples that populate your ontology.
- Compile a report of your project that contains the following:
- Brief narrative description of the topic that you pick to model.
- The data sources that you use as the basis of your modeling.
- The list of competency questions for your ontology.
- The list of key notions you obtained from the competency questions and one or more class diagrams that represent your ontology.
- List of all of axioms together with a brief English language explanation (one or two sentences) of each of the axiom. See chapter 1, section 1.6 for an example of how to explain the axioms -- the chapter explains axiom 1.1 - 1.49. Your explanation can be shorter and less verbose than the example.
- A snapshot of your data (RDF triples) in Turtle format.
- Present your work in a 15-minutes presentation on Thursday, December 8, 2016.
- Email your report, OWL file, and the RDF dump before the class on Thursday, December 8, 2016.