It would be interesting if the article was expanded in a number of directions: 1) How are the results from collaboration graphs used? Where do they tend to make non-obvious discoveries? 2) An alternate "weighting" of edges might have to do with which person was the more significant contributor. A little something to counter the trend in scientific journals of listing 40 people as all being the authors? 3) Placing elapsed time on edges. "Professor X wrote an article. Professor Y wrote a related article that cites X, but 25 years later. Then Professors Y and Z wrote six articles together."