The solutions should be ready for inspection by Thursday 21.2.2002 (midnight). After that, there won't be any new exercises for the following week (week 9) because we have lectures.
Remember that always, if you are in doubt what you should do, you can ask Reeta or send a message to our newsgroup!!
Read the paper E. Riloff:Automatically Constructing a Dictionary for Information Extraction Tasks and summarize its contents by answering the following questions:
What is the motivation of the AutoSlog system? Why has it been designed and implemented? What is the problem it tries to solve?
Consider the role of AutoSlog in an IE process. What is the input of AutoSlog? What is the output of AutoSlog?
Describe the basic "algorithm" of AutoSlog. How does it work? What does it do? (You can leave the details to the next exercise)
Describe more detailly how AutoSlog works using our sample terrorist data and the corresponding answer key templates.
Give (at least) 5 concept node definitions. You can present the definitions informally (not necessarily using the Lisp-style of the paper). Explain what your definitions mean (in such a way that also we can see that you have understood them...)