←back to thread

412 points thepuppet33r | 5 comments | | HN request time: 0.623s | source
1. zeroonetwothree ◴[] No.42175714[source]
Google Scholar is so good. I started doing research right when it came out and it was amazingly helpful. I can’t imagine how it was done before.
replies(4): >>42175801 #>>42176741 #>>42177679 #>>42186276 #
2. IshKebab ◴[] No.42175801[source]
There are alternatives, like Web of Knowledge. You basically need to be in a Uni for that though.
3. leephillips ◴[] No.42176741[source]
I would go to the library and pull volumes of Science Citation Index off the shelves. Yes, Google Scholar was a revolution.
4. dekhn ◴[] No.42177679[source]
I'd go to the card catalog (index), turn my question into a bag of words (tokenize), fetch all the cards matching each token (posting lists), drop cards which didn't include enough of the tokens (posting list intersection), ordering the cards by the number of tokens they matched (keyword match ranking), filter at some cutoff, and then reorder based on the h-index of the author (page rank). Then I would read each paper in order, following citations in a breadth-first manner.

(the above is a joke comparing old school library work to search engines circa 2000; I didn't actually do all those steps. I'd usually just find the most recent review article and read the papers it cited).

5. asdff ◴[] No.42186276[source]
I had an old boss who did it in the old analog way. He had a secretary handle his email and transcribing stuff he hand wrote. He had print subscriptions to a couple nature journals, science, and a couple research niche specific journals and he read them basically cover to cover. He'd attend conferences and had many collaborators who would send him papers from their own lab to opine on.

I actually respect this style a lot. There is a firehose of papers coming onto google scholar each day. You type in some keyword you get 500 hits. This cut that down substantially for him in a way where he never missed anything big (reading nature and science), kept up with what the field has been doing (reading the more niche specific journals and keeping up with the labs who put out this niche work), and seeing what was coming up in the pipeline from the conferences or what sort of research new grants were requesting. I'm not sure that scholar would have helped much.