A multithreaded crawler for ACM's digital library developed in Java. Note that ACM's digital library (dl.acm.org) will probably put this crawler to its blacklist pretty fast, and, thus, will prevent it from crawling the entire collaboration graph.
Just invoke the .jar-file without arguments; the help message will appear.
Once crawled, for yeah-experience, invoke one more time with --dump -option.