I see lot of relevant ads which are personalised to me on the Internet. But, I don't see the same thing happening with the Offline ads.

What it does

We are solving this problem by using the powerful data available on internet. This is B2B. Give us details what your company does, we show you the heat map of Tokyo which shows the place where yours ads would do better.

How we built it

Languages Used: Golang, Python, Javascript, C++ We divided Tokyo ( Area 2200 km2 ) into 926 blocks each of 4 Each block now has a geo-coordinate associated ( latitude and longitude ). We now traverse through all of these blocks use the geo data we have and get all the places within a certain radius. We now get the tags ( type of a place ) of each place. So, for each block we have count of each type of place present which we are going to use as data. From the client side, we ask on how do they describe their company and based on the options they give, we now use some algorithms to rank the blocks based on the data we have.

Challenges we ran into

  • While fetching the geo data, we have like infinite co-ordinates. So, we used depth first search algorithm to start from the middle of Tokyo and defined some limits and ran the algo to get the geo data.
  • We are team of all backend engineers, we needed some one to code the frontend part. learnt little of jquery overnight to build the frontend.

Accomplishments that we're proud of

  • We solved all the challenges we faced in to time.

What we learned

  • Having a frontend guy would have been better, we

What's next for smart-offline-ads

  • There's a lot of room for improvement. We can even further improve our algorithm. Dividing a place into much smaller blocks would definitely get more specified results.
  • We can even show the details of billboards present in the area, give the client details to contact the dealer directly which would make it much easier

Built With

Share this project: