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: Alternative Data to Google Search (Brainstorm)

1 点作者 quocble将近 6 年前
I was inspired by https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=20440079 thread regarding building a new search index. Technically it is feasible with CommonCrawl + cloud map reduce jobs. Clearly, you would be killed if you build google search, but I started thinking what kind of alternative data that you could build which could be valuable.<p>Here&#x27;s some of my ideas:<p>- Ecommerce products ranking based on backlinks.<p>- Competitor analysis - most linked pages, total backlinks<p>- Most common scripts installed - you can see trending in integrations such as react, fb, intercom, etc. ( already been done, i know )<p>Comments? Any more ideas?

1 comment

neotokio将近 6 年前
In 2015 &#x27;it will be killed off by google&#x27; sounded feasible in my mind. Today I use Duckduckgo for 90% of time and Google only for some very specific (mostly local) searches, Google seems to only hold &#x27;early starter and low moral privacy policy&#x27; advantage. They also seem to target typical &#x27;new Internet user&#x27; now more than ever, I don&#x27;t remember last time I found something useful or &#x27;smart&#x27; with Google.<p>That being said, let Google have general traffic search because it fails on everything else. I agree with your approach to find a generic niche, first question is - who&#x27;s end user? Will it be &#x27;I want to buy this item&#x27; (e-commerce)? Will it be &#x27;I want to find out industry specific information? (competitor). Will it be &#x27;I want to know how to use X&#x27; (your last one).<p>Each has its own problems. And, most likely, each already has a solution - it&#x27;s just probably a) too commercialized (e-commerce - comparison sites) b) biased (competitors - analytics engines used by whole industry) c) too broad<p>There is abundance of data and it interconnects heavily, there is more and more alternative data (social apps, images, maps etc) too. I think it would be nice to think outside of algorithmic approach (how do I find X in most optimal way) and start thinking about contextual approach (how do I find if X is right).<p>Uff, to the point - I think starting small, with something annoyingly hard or time consuming to find, would be best. For example, I always spend DAYS reading about basic house appliances before buying it and most of info I find is heavily commercialized therefore useless (affiliate programs). I don&#x27;t say &#x27;Do that!&#x27;, but there is definitely a need for niche-specific-search-algo&#x2F;aggregator.