Luca Dini wrote: > Hi, > we would like to use Longwell 2.4.2 as the preferred GUI for an > application of sentiment analysis which is based on rdfs descriptions > of the domain. The "backbone" storge for the application is a Lucene > 2.0 index. I have two questions: > 1) We noticed (in a previous version of Longwell, 1.0) that loading > the model in memory at startup (about 40,000 objects, rather simple > in structure) is much faster by using a class we wrote in the past > that creates the model out of the lucene index, rather than reading > it from an rdf files. Is there a way to integrate such a model > provider without hanging the source code?
no > 2) Longwell creates lucene indexes when loading the model and use > them for searching. In our case we already have the indexes and > search is more complex than the one currently implemented, in the > sense that only certain fields must be looked for, with certain > boosts etc. Is there a way to specialize textual search without, > again, modifying the source code? no Sorry but your requirements are not something that we forecasted at design time. Back compatible patches, as usual, are welcome though. -- Stefano Mazzocchi Digital Libraries Research Group Research Scientist Massachusetts Institute of Technology E25-131, 77 Massachusetts Ave skype: stefanomazzocchi Cambridge, MA 02139-4307, USA email: stefanom at mit . edu ------------------------------------------------------------------- _______________________________________________ General mailing list [email protected] http://simile.mit.edu/mailman/listinfo/general
