4 ms·
@jackcosgrove: Source? As I understand, the closest prior art to PageRank was HITS (as I called out earlier).
by Hydraulix989 7y ago
@jackcosgrove: Source? As I understand, the closest prior art to PageRank was HITS (as I called out earlier).
- jackcosgrove 7y agoI'm sorry, I was mistaken. It was Web of Science, not PubMed. Garfield, Garfield, Eugene. Citation indexing: Its theory and application in science, technology, and humanities. New York: Wiley, 1979, P. 1. http://garfield.library.upenn.edu/ci/chapter1.PDF http://garfield.library.upenn.edu/ci/chapter1.PDF During my investigation I found this article which claims an even earlier origin for this algorithm: https://m.phys.org/news/2010-02-google-pagerank-like-algorithm-dates.html https://m.phys.org/news/2010-02-google-pagerank-like-algorit...
- Hydraulix989 7y agoThese aren’t algorithms, they are just hand waving intuition that is “on the right track” but not quite there. You see similar “prior art” to Newtonian physics, for example, but there is still a huge leap from “oh, there might be something in these linkages” to “if we compute the stationary distribution of the Markov chain over the graph, we know the relative importance of each node.” Kleinberg was the first to algorithmically and rigorously codify this intuition.