search

moreymat / omw-graph

The Open Multilingual Wordnet in a graph database
MIT License
4 stars 0 forks source link

Merge parser and relation extraction #11

Closed zorgulle closed 10 years ago

zorgulle commented 10 years ago

By regrouping relation extraction and file parsing we can add other language easily but it is slower of 5 second

rhin0cer0s commented 10 years ago

It's okay for me if ouput files are like : rels-eng.csv words-eng.csv etc ...

Our process time is relatively low at this moment and we are not supposed to produce them a lot

Maybe it would be great if they are produced in a specific folder (csv_tmp for example) so it doesn't flood actual directories with files.

edit: furthermore if we are changing our sources ( partial or full OMW ) these timing are meaningless.

moreymat commented 10 years ago

@zorgulle why is it slower? You might want to use one or two of the profiling tools listed here to trace it.

moreymat commented 10 years ago

(5 seconds are definitely not worth it, but I had planned to give you this pointer eventually.)

zorgulle commented 10 years ago

By using Mr Bond's script we produce tab file, and csv files in 20 seconds. we realized that our parser is useless, because we don't read tab files anymore. do we have to keep the parser?

moreymat commented 10 years ago

@zorgulle it makes sense to delete the tab parser now. We certainly don't want dead code floating around the project.

moreymat commented 10 years ago

All cleaning has been done, we can close this.