馃懡 krixano

New Devlog Post: Search Engine Ranking Systems Are Being Left Unquestioned

gemini://auragem.space/devlog/20220722.gmi

2 years ago 路 馃憤 freezr

Links

gemini://auragem.space/devlog/20220722.gmi

Actions

馃憢 Join Station

5 Replies

馃懡 krixano

@freezr @haze Thanks! I'll read your post @haze, should be interesting to hear more opinions on this :)

@freezr I think NLP can help a lot, even with just rewriting queries. The problem with NLP is that you need to know about other languages and how they work to know how rewriting needs to be done, let alone how to parse even just "words" or units of meaning. 路 2 years ago

馃懡 haze

@krixano Wow nice post. Read it and I got the motivation to finish an old, related post that has been sitting on my capsule forever.

gemini://gemini.clehaxze.tw/gemlog/2022/07-22-search-engine-dilemma-bias-vs-accuracy.gmi 路 2 years ago

gemini://gemini.clehaxze.tw/gemlog/2022/07-22-search-engine-dilemma-bias-vs-accuracy.gmi

馃懡 freezr

I had time to read it! Nice piece! I really encourage you to experiment a lot! Not sure if NPL would help you but sounds more promising than the old alternatives... 路 2 years ago

馃懡 krixano

@mfoo2 Thanks! :D

I spent a lot of time writing the article, as well as a ton of time writing snippets of my thoughts on Station prior. I think I'm going to write a few more articles covering more problems with implementing Search Engines - basically the problems that I'm running into, as well as some of the problems me and @acidus have discussed recently here on Station. 路 2 years ago

馃懡 mfoo2

Thanks @krixano for your excellent article. It seems to me that these existing link based search engine algorithms mistake the syntax of the hypertext for its semantics. But a fundamental underlying problem is that there is no semantics of a link in HTML or gemtext so we cannot infer the type of relationship between any two pages. it's good to explore alternatives. 路 2 years ago