At the moment it is possible to select a BibTeX entry on a web page and via pressing the
postPublication button inserting it into BibSonomy. The next feature we will release next week allows to extract references from ACM or Citeseer without selecting a BibTeX entry. What we can already provide today is an interface for Scrapers and some helper classes which allow you to implement scrapers for other services. If you are interested in developing a BibSonomy-compliant scraper which we can include into the project, have a look into this JAR-file which contains the source code for the needed classes:
scraper-0.1.jar.