CS 8995

Tuesday March 20, 1pm

The deadline for stage 1 of the final project has been adjusted. 

Your gold standard data is still due at 4pm, Friday Mar 23. You must 
submit your gold standard data by this time, and it must be a unicode  
file that exactly follows the formatting guidelines provided. 

The rest of your stage 1 work is due by 4pm, Monday March 26. 

However, you must now evaluate your sentence alignment program not only 
using your gold standard data, but those of ALL the other teams. Report
those results in a table in the write up you will submit on Monday. I 
will post all of the gold standard data on the class web page before 6pm  
Friday. 

Your table should look like this (assume that my evaluation metric
assigns scores from 0 to 1000, your metric may follow whatever scale
you like, but it should ultimately report a single value). 

Assume that this is for team PUEBLA. 

CRUZ AZUL   900
TOLUCA      800
MORELIA     990
UNAM        200
PACHUCA     700
TIGRES      400
PUEBLA      800

This table tells me that the PUEBLA sentence alignment program scored
900 on the CRUZ AZUL data, 800 on the TOLUCA data, 990 on the MORELIA
data, ..., and finally 800 on your own data.

Once you have turned in stage 1 of the project, I will post your
evaluation programs. You will then evaluate the results of your
sentence alignment program using each of the different evaluation
programs. This must be completed by Weds Mar 28. Further details
will be provided on Monday. 

Since we are sharing data and code, it is crucial that your data
follow the established format, and that your code run as I describe
on the class web page writeup. Follow the naming conventions and
parameter usages given exactly! Make certain that your code runs
on the csdev platforms exactly as requested. 

Good luck!