In my opinion Apache OpenNLP has demonstrated a nice understanding of the Apache way, I think one thing to keep eyes on is the so called 'bus factor' [1] as it seems Jörn is far the most active committer [2]; it'd be good if there is a more balanced commit rate between all the committers. However, apart from that, I'm +1 for graduation to TLP. I'd advice to take a look at the 'guide to successful graduation' [3] in order to understand what the board and Incubator PMC expect from a project graduating to TLP. Tommaso
[1] : http://en.wikipedia.org/wiki/Bus_factor [2] : http://svnsearch.org/svnsearch/repos/ASF/search?path=%2Fincubator%2Fopennlp [3] : http://incubator.apache.org/guides/graduation.html#toplevel p.s.: WRT Lucene/Solr integration I've given a talk at latest LuceneEurocon about 'Natural language search in Solr' ( http://www.lucidimagination.com/devzone/events/conferences/ApacheLuceneEurocon2011_presentations#tommaso_teofill) and used OpenNLP inside UIMA to demonstrate how NLP can help on improving recall/precision. 2011/11/22 Jörn Kottmann <[email protected]> > +1 to become a TLP > > Jörn > > > On 11/22/11 8:46 PM, [email protected] wrote: > >> +1 TLP >> >> On Tue, Nov 22, 2011 at 4:07 PM, Boris Galitsky<[email protected]** >> >wrote: >> >> >>> >>> OpenNLP sees good adaption/integration by other Apache projects, such >>>>> >>>> as >>> >>>> Stanbol, UIMA, >>>>> Lucene/Solr (via UIMA and direct integration is planned) and Clerezza. >>>>> >>>> All >>> >>>> these collaborations >>>>> are a good advertisement for us and will attract more users over time. >>>>> >>>> I was proposing at Lucene/Solr meetups here in Bay Area that there is a >>> need for 'linguistic' analyzer which filters out high TF*IDF but >>> irrelevant >>> answers based on parse-features.So I will be further advocating OpenNLP >>> and >>> syntactic similarity-based SOLR request handler. >>> RegardsBoris >>> >>> >
