For how long do research papers remain relevant for us?

[Marcia]

I've been wading through some research papers, primarily 4 of them, and a lot of what I've seen for myself over time, plus read as posted and experienced by other members here at the board, is covered in those papers. A lot of what's in those gives deep insights into factors that can be taken into consideration for search engine rankings as well as possibly be the cause of penalties.

What I'm wondering about is at what point the information we read and extract from those academic papers may be outdated, or if they remain important for us indefinitely because of the conceptual principles they divulge.

Anyone have any thoughts or experiences? Are they worth perusing and studying again for current or future value to us?

[jeremy goodrich]

Some are 'timeless' because they contain nuggets that, imho, won't change over time.

Eg, the classic HillTop algorithm, or the infamous PageRank paper with their new famous equation, etc.

There are technical details there that 'fit' with my own analysis of Google's results, as well as the results of other engines.

Some stuff, though, is bound to be 'out of date' for example the whole term vector database idea as a stand alone algo, these days, it's much more likely that the algos are all going to revolve around some form of Kleinberg's hubs & authorities.

[vitaplease]

I find the research papers the most valuable thing around (next to WebmasterWorld).

Never mind if things mentioned are way off computationally, they voice mechanisms, web-polution, new ranking possibilities and short-comings of existing algorithms etc.

The pity is, that whilst many mention new published research papers, few comment on them, but then some of the stuff is so good, it would be a pity to post about them, see ya at the pubcon :)

[tedster]

The papers I've studied give me a change of viewpoint that is extremely valuable. All of a sudden I'm looking at the angle of how to deliver relevant answers to queries instead of how to make my pages come out on top. It's a very different mindset - and it disabuses me of a lot of the "looking for tricks" that I used to be so keen on.

From that point of view, historical papers retain value - even papers from long before the web was born. Back then Library Science and Information Science were extremely esoteric studies followed by a few rare people who spoke about mysteries I couldn't be bothered with. What do you know - they were really on to something important.

I've always had a soft spot when it comes to "learning for learning's sake". And a lot of things I've picked up along the way ended up having a practical value decades later - even though that wasn't why I read them.

So, as I have the time, I read what I can get my hands on. Who knows what I'll uncover.

[mona]

>Are they worth perusing and studying again for current or future value to us?
I think they are. The first time I read PageRank paper was well over a year ago and it only confused me. I just finished reading it again and I'm amazed by what I learned.

>From that point of view, historical papers retain value
That's part of what I found so fascinating on the second read, the historical value of what's written.

To test the utility of PageRank for search, we built a web search engine called Google.

I'm embarrassed to admit I didn't even know Google was made for PageRank, I always thought it was the other way around...