Results of Hackfest

Hackfest was held in Hamilton Public, central branch.  It was a good opportunity to meet other System Librarians and to talk code.  I choose to work with others on the loosely based idea of analysing OPAC log files and turing them into something that could help with collections decisions or to get a feeling for the most popular searches are. 

We had the good fortune of having a few days worth of usage data from UPEI's Evergreen catalogue.  The predictable URL syntax and the good level of detail in log files allows anyone to scoop out lots of meta data about searches, results, and records examined. So the plan was to grep out some useful log files that surrounded key word searches and do something neat with them.  The something neat, Solr.   As it turns out there is this great java based enterprise search platform that when given XML schemas and datasets will crank out xml search results.   What could be better?  After a full day of mushing data the end result was pretty neat.  A 'Most Popular Search Terms' list that is fed with data from the catalogue.  Now the usage data was confidential, so I can't really post the exact results.

Now imagine utilizing Solr to create some sort of catalogue that has all of these great search options and data formatting.  Turns out that already exists and its called VUFind. In my opinion this is a top contender for faceted, open source, neato catalogue platforms.   The only drawback is Java, but the ends might justify the means in this case.