Collaborative filtering in Python or Ruby and MongoDB.

Web, Mobile & Software Dev Scripts & Utilities Posted 1 year ago

Fixed Price

Delivery by April 7, 2014




We have customer orders and quotations stored in MongoDB. You need to write a script in ruby or python to implement collaborative filtering. If it is Ruby, then you may want to account for RoR. If it is in Python, then feel free to utilize Numpy and Pandas.

Both customer order and quotation have the same structure. { _id:ObjectId, partner_id: ObjectId, content: [{ product_id: quantity}] }. The two collections to be accessed and need to be combined and analyzed as one.

Data has to be analyzed using collaborative filtering techniques in various dimensions. Specifically:

1. Product_1 is similar to Product_2 based on customer orders. That is "when this product was bought these products were also bought" I.e. A vector of a product describes customer orders, [quantity in order 1, quantity in order 2, quantity in order 3, etc]
2. Product_1 is similar to Product_2 based on partner_id. That is "Those who bought this product also bought" I.e. A vector of a product describes partner ids.
3. Partner_1 is similar to Partner_2 based on partner_id. I.e. A vector of partnrt_id describes product_id.
4. Your ideas will be awarded separately :)

You will be provided with sample data in Mongo dump-json format.

The result need to be saved to MongoDB in an easily accessible format. For example, { product_ids: [ObjectId_1, Object_Id_2], similarity: float }

Skills Required:

Client Activity on this Job

Last Viewed: 1 year ago

Applicants: 6

Hired: 1

About the Client

(5.00) 22 reviews

Moscow 04:17 AM

39 Jobs Posted
85% Hire Rate, 1 Open Job

Over $10,000 Total Spent
42 Hires, 5 Active

$22.20/hr Avg Hourly Rate Paid
644 Hours

Member Since Mar 8, 2013