PolNet is a WordNet like lexical data base built from scratch according to the "merge model" methodology. Its design started in 2006 and continues. The resource development procedure is based on the exploration of good traditional dictionaries of Polish and language corpora investigations (IPI PAN Corpus and domain/application oriented corpora). The PolNet development was organized in an incremental way, starting with general and frequently used vocabulary. We selected the most frequent words found in a reference corpus of Polish language with however one important exception made for methodological reasons. The reason was that we assumed possibly early validation of the resource in a real-size application for which an application-complete vocabulary was necessary.