Skip to content

Ask HN: Google web corpus + big data analysis cluster?

3 pointspasteurquadrant1 comment
On HN

Thinking about starting a developer friendly service that frequently crawls the web at massive scale + a big data analysis cluster (such as Storm) with support for a variety of languages.

Obviously walled gardens like Facebook and Twitter would be off limits at the beginning, but if the service gains traction, then it's possible companies would want to be crawled.

I was trying to analyze the common crawl recently, and the process of getting set up is non trivial. This service would allow more people to more easily analyze the web.

Would welcome any feedback.

Comments

Some quick questions:

1) How frequently would you crawl?

2) What are you defining as a 'massive scale'? It'll probably grow i'm sure but what is a reasonable range you're working with?

3) Which languages will you be supporting?

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.