Office hours: Tuesdays 12:30 to 1:30 (except for Tuesday March 16)
Wednesdays 1-2pm
No office hours during spring break!
Also, if you are unable to make these times, please send me email. I am happy to arrange to talk at other times.
TA: Tim Greenfield
email: greenfit@cs.colorado.edu
2) Learn about several state-of-the-art AI approaches for supporting
information retrieval in web-based information spaces - a knowledge modeling
approach and an alternative learning approach (Latent Semantic Analysis)
will be covered in class
3) Understand what the key problems and challenges are for supporting
effective information retrieval
4) Be able to identify and critique the claims made about these different
systems and approaches.
5) Be able to compare and contrast the strengths and weaknesses of these different approaches with respect to different enterprise settings.
Duration: 8 class sessions
Block 3 Assessment: 30% of total course grade (30 points total)
up to10 points: participation in take-home and in-class activities
up to 20 points: project plus report
A Knowledge-based News Server Supporting Ontology-Driven Story Enrichment
and Knowledge Retrieval
by John Domingue and Enrico Motta
To appear in: The 11th European Workshop on Knowledge Acquisition,
Modelling, and Management (EKAW '99).
PDF Available at: http://kmi.open.ac.uk/~enrico/
NEW FIGURE 5: Some of you will notice that the event hierarchy (Figure 5) in the Domingue and Motta article is unreadable. John Domingue has kindly sent us an enlarged readable version of the Planet-Onto event hierarchy. John has also sent us pictures of the kmi-member, kmi-technology, and organization ontologies.
An Introduction to Latent Semantic Analysis
by Thomas Landauer, Peter Foltz, and Darrell Laham
In: Discourse Processes, 25, 259-284, 1998.
PDF Available at: http://lsa.colorado.edu/papers.html
LSA website: http://lsa.colorado.edu/
Frames and Frame Systems
in your text by Russel and Norvig
pages 316-323
Agents that Reduce Work and Information Overload
by Patti Maes
Communications of the ACM, Vol. 37, No. 7 (July), 1994, pp. 31- 40.
Symbolic Modeling in Practice
by Hermann Kaindl and John Carroll
Communications of the ACM, Vol. 42, No.1 (January), 1999, pp. 28- 30.
The Cognitive Ergonomics ofİ Knowledge-Based Design Support Environments
by Tamara Sumner, Nathalie Bonnardel, Benedikte Harstad
Conference on Computer-Human Interaction (CHI '97), Atlanta
Available at: http://kmi.open.ac.uk/techreports/papers/kmi-tr-38.pdf
The Vocabulary Problem in Human-System Communication
by G. W. Furnas, T. K. Landauer, L. M. Gomez, and S. T. Dumais
Communications of the ACM, Vol. 30, No.11 (November), 1987, pp. 964-971.
Recommender Systems
by Paul Resnick and Hal R. Varian
Communications of the ACM, Vol. 40, No.3 (March), 1997, pp. 56 - 58.
Tuesday, April 6 - More on LSA
Gerry Stahl, one of the creators of the LSA Summary Street program, will come to class to answer your questions about LSA, including:
Mike Wright from Unidata/UCAR will come to class to talk about why doing user profiling on the world wide web is hard. Mike will describe his work in this area on developing user constituencies (profiles of groups of users), rather the profiling or modeling individual users. You can read about Mike's project on User Constituencies in this shortish (around 8 pages) article.
A 'summary report' is to be prepared and emailed to Tim and Tammy by noon on Wednesday, March 31. For the subject line of your email, use "cs3202 - LSA summary". This report should take you about 30 minutes. On Thursday, April 1 we will use these summaries as input for the Essay Anlayzer software and you will have the class period to further refine your summaries based on feedback from the Essay Analyzer. So bring a copy of your summary with you to class on Thursday. Provide the following information for your initial summary:
Your Name!!!!
Title of paper
Author(s) name(s)
Three paragraphs, one for each of the three parts of the paper. Each paragraph should be between 150-200 words, for a total length of 450-600 words. We need the summary structured according to parts so we can feed it into the Essay Analyzer software.
Three questions that can serve as the basis
for class discussion. Some sample question templates include:
How does ...
relate to ...?
How are ...
and ... similar or different?
Is ... consistent
with ...?
Does ... support
...?
What is the
meaning of ...?
How would one
use ... to ...?
How does ...
affect ...?
Does it seem
reasonable to claim ...?
Does ... imply
...?
A brief sketch (one paragraph) of the issues
concerning one of the questions you raise. Essentially this is an expansion
of the question providing a little context about it.
Knowledge Modeling Assignment related to the Domingue & Motta Article
The Domingue and Motta paper described an operational system based on the existence of a well-designed ontology. In this assignment, you will gain some first hand experience creating an ontology. This assignment should take you around 45 minutes.
The task: You are to create an ontology that describes the key concepts and events contained in your article assignment from Project 1.
Represent your ontology using a similar representation to that used in the Domingue & Motta article; i.e., a graphical frames representation. Specifically, represent concepts (classes, instances, relations) using nodes that are connected by "is-a" links. Elaborate on the definition of concepts by defining their slots and slot-values as appropriate.
To hand in: Your annotated article and ontology.
Some hints:
SUMMARY REPORT for Domingue & Motta Article
A brief required 'summary report' is to be prepared and emailed to Tim and Tammy by noon on Monday, March 15. You should not spend more than 20 minutes preparing these reports (unless you just want to!). Provide the following information for this paper:
Your Name!!!!
Title of paper
Author(s) name(s)
One paragraph summary
Three questions that can serve as the basis
for class discussion. Some sample question templates include:
How does ...
relate to ...?
How are ...
and ... similar or different?
Is ... consistent
with ...?
Does ... support
...?
What is the
meaning of ...?
How would one
use ... to ...?
How does ...
affect ...?
Does it seem
reasonable to claim ...?
Does ... imply
...?
A brief sketch (one paragraph) of the issues
concerning one of the questions you raise. Essentially this is an expansion
of the question providing a little context about it.
Block 3 Projects and Project Proposals
Option 1: Continue with AS project, hand in final report
- same groups (or subset)
- everyone gets same grade
- Must have a clear ADVANCE
- proposal guidelines: one paragraph - what I did in project 2; one paragraph on the anticipated advance
Option 2: 2000 word report on "AI and the Web"
Examples:
- focus on a technique or technology; e.g. recommender systems or ontologies
- focus on a specific application of use setting; e.g., My Yahoo, Amazon, other..
- focus on an application area; e.g. digital libraries, e-journals
- other ideas????
- proposal guidelines: one paragraph - what is your focus; one paragraph on anticipated approach
