3 ms·
When google got started, they indexed ~25 million web pages. Then they grew together with the web, compounding over ~30 years. Today, it’s totally possible for
by wenbin 5y ago
When google got started, they indexed ~25 million web pages. Then they grew together with the web, compounding over ~30 years.
Today, it’s totally possible for a single person or a small team to build a domain specific search engine that indexes 10s or even 100s of million documents.
Building the index v0.1 that kind of working is not hard. But maintaining the index to handle countless edge cases is tedious and need non-stop investment (eg, infrastructure + paying salary for talents)
Also, a search engine is more than “keyword matching”. You need to do search result ranking. And it gets exponentially difficult to do as # of indexed documents increases.
- a9h74j 5y agoEmpirical question: Can any kind of cluster analysis result in identifying robust selected subject-domains/content-areas for search specialization, still allowing for good coverage with a set of these? This could fight what you identify as an exponentially difficult problem. It could also allow for selective marketing and mind-share.