Google Code-in

apertium-anaphora is a module that was recently added to the apertium pipeline. Currently it is used by a branch of the apertium-eng-spa pair.

You need to take a corpus of 10000 Spanish sentences. Document the time it takes for apertium to translate to English through the default pipeline. Then, use the pipeline with apertium-anaphora (added right before apertium-transfer in modes.xml). Again time the translation and report both times.

The second thing you need to complete this task is time the anaphora module individually. To do this, you need to manually run each module in the pipeline till the lrx module. Take the output of this module and put it in a text file. Now, simply give this file as input to apertium-anaphora and time it.

Helpful links:

• Documentation: You can understand the apertium-anaphora module here.
• Repo: See the README to understand how to run apertium-anaphora.