Forum Moderators: open
In recent years backlinks have slowly gained more and more authority to the point where they have made all other factors almost obsolete. However, the problem with backlinks is that they reward popularity, rather than relevance, which is to my mind the main purpose of a search engine.
The theory of LSI would surely make body content less manipulable and reduce the dependance upon backlinks and PR; but is this likely to happen?
Does anyone believe that LSI now is, or will become in the near future a pivitol factor in deciding rankings?
I'd appreciate any thoughts.
It's the holy grail of search engines, you can bet that all the big players have people locked in rooms working on this, they just haven't cracked the scalability issues.
I have personally used LSI for a project where 268,000 documents were analysed. The amount of calculations required to produce sensible data from that was astounding (read days). Imagine scaling that up to a web sized database!
There is also the quality of data going in, search engines clearly have an issue identifying spam sites so it could be that such sites find their way into the dataset, this would upset the calculations (but probably not significantly enough).
My guess is that there is an alterior motive to Google Scholar/Books - it's to create a massive database for LSI research, data that has been edited and is less likely to be spammy. The problem there is that it still would not give as wide a set of data as the web, but losing 10% of relevance in exchange for better results on the other 90% has never bothered Google before, it's collateral damage.
Whatever happens it's going to be interesting.
[webmasterworld.com...]
[webmasterworld.com...]
