Very cool. What's the software stack behind this? How are you performing searches across all this very diverse data? For example, you must be doing some sort of text-analysis to retrieve the mileage property out of the text, no?
Craigslist is a ton of NLP to figure out common ways to phrase mileage etc... also catch all the mispellings.
We're also experimenting with some structured parsing for the universal dealership parser we alluded to - this means trying to figure out which part of the page has a listing so we dont accidentally report the mileage from a warranty or similar listings lower down the page... turns out to be a fairly tricky problem but its also a ton of fun to work on! Will keep you posted if we start having a ton of success automagically crawling the interweb. Thanks!
Comments
Very cool. What's the software stack behind this? How are you performing searches across all this very diverse data? For example, you must be doing some sort of text-analysis to retrieve the mileage property out of the text, no?
Craigslist is a ton of NLP to figure out common ways to phrase mileage etc... also catch all the mispellings.
We're also experimenting with some structured parsing for the universal dealership parser we alluded to - this means trying to figure out which part of the page has a listing so we dont accidentally report the mileage from a warranty or similar listings lower down the page... turns out to be a fairly tricky problem but its also a ton of fun to work on! Will keep you posted if we start having a ton of success automagically crawling the interweb. Thanks!