UMBC CMSC 491/691-I Fall 2002 Home  |  News  |  Syllabus  |  Project   ]
Last updated: 17 October 2002

Homework 4

Assignment: Develop three search topics for the umbc-crawl collection.

Goal: To turn umbc-crawl from a web crawl into a usable test collection.

Due date:Tuesday, October 22, 2002.

Description

To create a topic, first think of an area of interest within the domain of umbc-crawl. Then, using one or more search engines, search for pages relevant to your topic using one or two queries. Try not to build topics that are too easy, or that have too much relevant information available, or that have too little. In your initial searches, you shouldn't need to go past the first or second page of hits to decide if the topic is too easy or not. Also, don't feel that only "univerisity-genre" topics have information in the crawl... some people (and departments) have very eclectic web pages!

Some search engines you might try:

After you have searched for your topic, and learned a bit more about what you want to consider relevant and irrelevant, compose your topic statement using the following TREC-style format:

     <top>
     <title> Short title
     <by> My name here (me@umbc.edu)

     <desc> A one sentence description of what I am looking for,
     containing the title words and other important keywords.

     <narr> A short paragraph spelling out my information need.  The
     narrative specifically indicates what kinds of information is relevant
     and what is not.
     </top>

You will create three such search topics. Be creative; if everyone searches for the same thing, we'll have a very boring set of topics!

What to turn in

First, put all three of your search topics into a single ASCII test file, and submit this file to me via the BlackBoard digital drop box. PLEASE do not use Postscript, PDF, HTML, Microsoft Word, or anything else except plain text. Otherwise, I will ask you to resubmit.

Second, bring three copies of each of your topics, on separate pieces of paper, to class on Tuesday. We will swap the topics in class on Tuesday in order to set the stage for Homework 5.