TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: Can you borrow Googles search algorithm?

10 点作者 brittpart_超过 4 年前
Or do I have to start from scratch? Maybe relating to NLP.

6 条评论

quickthrower2超过 4 年前
More prosaic but if you have a blog etc. you can add a Google programmable search engine: <a href="https:&#x2F;&#x2F;programmablesearchengine.google.com&#x2F;about&#x2F;" rel="nofollow">https:&#x2F;&#x2F;programmablesearchengine.google.com&#x2F;about&#x2F;</a>
评论 #25440244 未加载
HakeHayashi超过 4 年前
Mm... PageRank is the basis, but then recent publications suggest they use a multitude of factors (probably 7-10+?) to evaluate sites. PageRank is still more-or-less the biggest chunk, but since people learned how to game it, they&#x27;ve had to likewise step up their internal game on ranking. What you&#x27;re requesting probably comes off as a trade secret, but you could probably get reasonable results using a PageRank inspired hybrid
评论 #25415658 未加载
softwaredoug超过 4 年前
What are you trying to achieve? Search your website&#x27;s blog posts? Create an e-commerce search engine? Job search engine? Web search engine? Just learn about how search tech works?<p>All of these are radically different domains, some of which requiring intensive NLP, others requiring other domains...
logicslave超过 4 年前
Funny thing I was thinking about recently...why not reverse engineer it? Run the top million most common queries, or maybe top billion, snapshot the top 100 results, use that to train your model. With cheap enough compute, can google be reverse engineered?
评论 #25414685 未加载
throwaway888abc超过 4 年前
<a href="https:&#x2F;&#x2F;developers.google.com&#x2F;custom-search" rel="nofollow">https:&#x2F;&#x2F;developers.google.com&#x2F;custom-search</a>
评论 #25418626 未加载
ktpsns超过 4 年前
I would recommend to start with a mature search engine such as Lucene or ElasticSearch.
评论 #25418621 未加载