I am trying to build an application that can take a record of a car from different websites, compare it to data i have in a CSV file and return me the matching row.

Each website will present and structure the data in different ways. This is so I can know that the car I am comparing across websites are in fact the same make and model of car.

Data in my CSV file is highly structured and includes fields such as number of doors, cylinders in the engine, transmission speed etc.

I cannot trust that the data will be presented in the same way. (eg. VII vs 7, 1.4T vs 1.4 Turbo) Ideally I would be able to import a file with the unstructured data somewhere, have some kind of ML algorithm review and clean the data and spit out a version of the input with all records standardised.

I feel it is similar to the old programs that would fill in metadata for songs back in the day but given that so much data is provided in the input albeit not in a standard format we should be able to do this right? I am very new to world of Data Science/ML and know my way around python and have played with MS Cognitive Services. I am not sure what to even search for to solve this.

Hopefully someone can help.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.