All geek questions in one place

Amazon as an e-commerce site and recommendation system

I plan to implement a basic recommendation system that uses Facebook Connect or a similar API for social networks to connect a user profile, conduct tag-based analysis and use the results to create recommendations on items on my e-commerce site (it works similar to Amazon).

I believe that I need to divide the parts into the following:

  • Getting social networks through the API. (Indeed, this allows the user)

  • Analyze this data and generate Tokesh.

  • Using information markers, make a recommendation element on my e-commerce site.

EG: I am a fan of The Strokes on my Facebook account, the system analyzes this and recommends me the The Strokes Live CD.

For any part (data selection, tag-based recommendations ...), which algorithm and method would you recommend / use?

+4
source share
2 answers

If I were you, I would definitely take a look at the algorithms and articles published during the Netflix contest. See the forum and sites of the best teams.

+2
source

Good practical books on these types of algorithms:

+8
source

Source: https://habr.com/ru/post/1307231/

More articles:


All Articles