Web search results on par with results from Google?

I've tried DuckDuckGo (https://duck.com) , Brave Search (search.brave.com) and the others, I've tried SearXNG (https://searx.space), and I just can't get search results as good as when go to google.com. 4get and SearXNG's Google proxies get rate limited, as well as Whoogle, but even Startpage (https://startpage.com) doesn't give me results as good as when going to google.com, and that is a direct proxy service for Google's search. What am I to do? Also how does Google get search results so good?

>>25711 4get: https://git.lolcat.ca/lolcat/4get 4get.ca (main instance) Whoogle: https://github.com/benbusby/whoogle-search

wait google search results are good ?

>>25713 Honestly better than whatever else is out there imo. I don't primarily use Google, but it is the second best choice when my first fails me.

>>25713 Yes they are very good, it's almost as if they can read my mind sometimes. >>25714 First choice?

>>25715 >First choice? oh yeah mb. its duckduckgo. it is serviceable enough. though i might try startpage.

kagi is much better than Google. worth the price to me

>>25711 Why has everybody urgently started looking for google alternative? Has anything happened?

Google is not on par with Google. They broke their own search engine for ad revenue a few years ago. Use the EU servers when using Startpage. I think US servers possibly use Bing while the EU ones use Google. I may be wrong though but compared side by side I get different results for the same search when using US vs EU with all other settings the same. Originally Ixquick used Google and Startpage used Bing but they merged and I think the EU servers are what Ixquick merged into. That or it is just different results from region based results slipping in. https://eu.startpage.com/ I find Right Dao fairly good in results even if very limited it will compare to modern Google at least which doesnt really give you full access to results anymore anyway, The site runs from New York though and you kind of have to just trust their word they arent collecting your data but you're doing the same thing when you use even open source metasearches because you dont know what they really are running on their server. https://rightdao.com/ Yandex I find sometimes does as good as Google did before they broke themselves depending what you look for. But you also commonly have random Russian sites work their way into results even if region and language is set elsewhere. You can use it via various metasearch engines but really if you aren't Russian you actually likely fine allowing the Russian government to possibly be collecting your data since they wont be handing it over to enemy states you'd more have to worry about if the company that owns Yandex is selling your data. So your choice on if using it directly or via a metasearch. https://www.yandex.com/ Yahoo Japan seems to turn up less SEO content. Japan has decent data protection laws on par with the EU though unfortunately. > Since 2010, Yahoo! Japan's search engine has been based on Google's search technology. In exchange, Google receives user activity data from Yahoo! Japan's various products. > Yahoo Japan's services are not available in the European Economic Area and the UK since 6 April 2022, due to "excessive regulatory burden". Also results tend to be full of Japanese sites mixed in but you still get English sites too so it's not a big deal. American Yahoo on other-hand uses Bing for it's backend. https://search.yahoo.co.jp/ MetaGer used to be good aggregated metasearch but they recently made it more difficult to use and more a pay setup when before it didnt have all this token thing or limitations on what search engines you can use in combination. https://metager.org/ https://metager.de/ Stract is useful in filtering searches but results are limited depending what you want to find https://stract.com/ If you're looking for academic stuff maybe try WolframAlpha. I find it a bit difficult to figure out how to use well though so there is a learning curve. https://www.wolframalpha.com/ Naver from South Korea doesn't turn up any English language sites. Likewise the Chinese Baidu fails at this too. But if you need something Korea or China site specific then you can use these. Mwmbl the results aren't good right now unfortunately but it is user curated so it can be improved. Wiby and Marginalia are a very specific niche of finding sites that are more bare bones or from the oldweb so their use is very limited.

>Yandex Don't use Yandex. It is under Putin. In fact Putin has even visited its headquarters personally https://www.bbc.com/russian/news-41350939 during the period when Volozh used to be its CEO. After the full-scale invasion he has split the company into 2 parts, fully gifted one of the parts to the regime (in fact it is more correctly to say that the regime has given the another part to Volozh, and we can never be sure what has Volozh granted to the regime in exchange, you know, why to let go a part, if one can keep whole? So if I was an immigration authority, I wouldn't have given Volozh even a simple tourist's visa), and now pretends that he is clean. The photo is from www.kremlin.ru and under CC-BY 4.0. P.S. Ilya Segalovich, R.I.P. Yandex was much better when he was alive.

>>25723 vladmir putin never called me an incel

>>25711 >Google proxies get rate limited >>25721 I don't know much about this stuff, but this must be related.

Use searx wordswordsqordswirds

>>25723 > Don't use Yandex. It is under Putin. In fact Putin has even visited its headquarters personally Yeah and…? I explained that already it doesn't mean anything if you're from a country against Russia unless you're a glowie because you are in no way under the thumb of the Russian government. You only have to worry about the company itself selling data to the government and companies where you live if anywhere if the data of your searches is sensitive enough. If you're from Russia then yes don't use Yandex. If you're from Ukraine and are a soldier then dont use Yandex. If you're a CIA agent then don...wait why is a CIA glowie here.

>>25729 There are possible concerns of the data being collected from citizens of a foreign country to be used in campaigns directed by foreign entities for propaganda purposes. Honestly, if Yandex is a serious concern to the US, it would have been put on a watchlist years ago. I think using Yandex is fine. It is also fine if you don't want to use anything tied to current Russia. >>25722 >WolframAlpha. I find it a bit difficult to figure out how to use well though so there is a learning curve. So far, I only use wolframAlpha as an extremely overkill calculator. Helpful with finances (Be careful what you feed it. just in case). It is also a service (from many more) I considered using for a DIY personal assistant project, before GPT came out. The project fell flat because I was stressing about college, but i'm slightly curious to look into it again.

To be honest, Yandex search results quality has been soykaf, compared to Google, the last time I have used it.

bit unrelated but has anyone tried yacy? it's a distributed search engine not reliant on a central server, and you can run your own node. https://yacy.net/

>>25733 > There are possible concerns of the data being collected from citizens of a foreign country to be used in campaigns directed by foreign entities for propaganda purposes. Honestly, if Yandex is a serious concern to the US, it would have been put on a watchlist years ago. I think using Yandex is fine. It is also fine if you don't want to use anything tied to current Russia. I would think they would only bother with high profile people. If you're not famous or a politician no one cares to use you that way. What are they going to do make a graph of average citizens? lol I'm just saying I haven't heard of anything like that done before on ordinary foreign citizens. Ordinary foreigners only get involved in cases of being political hostages when in a foreign country which is outside the scope of search engine use. > So far, I only use wolframAlpha as an extremely overkill calculator. Helpful with finances (Be careful what you feed it. just in case). It is also a service (from many more) I considered using for a DIY personal assistant project, before GPT came out. The project fell flat because I was stressing about college, but i'm slightly curious to look into it again. If I recall correctly Amazon Alexa uses WolframAlpha but I may be thinking of some open source project I cant recall instead. >>25734 It's a mixed bag. Some searches turn up good results some turn up useless results. >>25740 Everyone I've seen who used it say the results aren't good yet.

>>25740 I used it a little some time ago. Utter horseshit in terms of search results. The network was too small, and too slow, and full of garbage, redundant entries, broken links, etc etc. Even when pages were there the search engine ranking system needs quite a bit of work, and this on a very small data set compared to what the dataset needs to be! The scale of the web that one must crawl in order to make a general engine is extremely large and yacy is not well equipped for such an undertaking. Certainly fine for very limited scope engines like your own site or perhaps even niche topic where crawling adjacent domains is desirable.

>>25741 >>25742 that's a shame but it's a good concept, still might try it out though

>>25742 >>25746 Is Seeks any better? https://beniz.github.io/seeks/ https://github.com/beniz/seeks

I wonder if we can summarize the content of documents using cheap local LLMs or API for OpenAI/Google and very-high-quality prompts for them, then vectorize the summaries, and then make YaCy into a distributed vector index rather than just a keyword index.

>>25751 No get info from people not machines then you can properly vet it instead of being fed hallucinations.