Discussion Class 7

1
Discussion Class 7

• User Requirements

2
Course Administration

• No Office Hours Thursday, October 17 or Tuesday
  October 22.
• Guest Lecture Tuesday, October 22. Carl Lagoze,
  Distributed information retrieval
• Midterm Examination Wednesday, October 23,
  Upson B17, 730 to 900. A sample examination
  and a discussion of the answers is on the web
  site.
• The midterm and final examinations aim to reward
  regular attendance in class (including guest
  lectures) and careful reading for the discussion
  classes.

3
Discussion Classes
Format Question Ask a member of the class to
answer Provide opportunity for others to
comment When answering Give your name. Make
sure that the TA hears it. Stand up Speak
clearly so that all the class can hear
4
Question 1 Distributed Information Retrieval
"The more general problems are locating the best
databases to search in a distributed environment
that may contain hundreds or even thousands of
databases, and merging the results that come back
from the distributed search." (a) What is the
problem? (b) Is this a problem for web search
services such as Infoseek or Lycos? How
do they handle it?
5
Question 2 Efficiency
(a) As computers get more powerful, the need for
efficiency often diminishes. What is special
about information retrieval? (b) What aspects
of efficiency would you consider especially
critical in storage, processing, and memory
utilization? (c) What has changed in the seven
years since this paper was written? What about
the next seven years?
6
Question 3 Relevance Feedback
"These factors mean that traditional feedback
techniques can be unpredictable in operational
settings." (a) What assumptions does relevance
feedback make about user behavior? (b) How is
practical experience different? (c) Is this a
fundamental problem?
7
Question 4 Integrated Solutions

"The most important problem from the point of
view of companies using and selling text-based
systems is integration with other systems." (a)
Give examples of integration of information
retrieval with other systems. (b) What impact
has the web had on this situation? (c) What are
the problems in integrating information retrieval
into database systems?
8
Question 5 Magic

"One of the major causes of failures in IR
systems is vocabulary mismatch." (a) What has a
thesaurus to do with this statement? (b) What
has latent semantic indexing to do with this
statement? (c) Are there problems with "magic"?
9
Question 6 Precision and Recall
Croft suggests that precision and recall may not
be the only criteria in the usefulness of the
results from information retrieval. (a) What
else does he consider important? (b) What does
this have to do with stemming?
10
Question 7 Multimedia retrieval
(a) What is the problem of multimedia
retrieval? (b) Why does Croft put it so low down
his list (8)?