Recommedation Engine – [Item to item collaborative filtering]

Recommending Engine which leverage fresh suggestions from massive items from Amazon, Google. All these run by one gear called Machine Learning lets split down the issues and possible solutions as below,

Problem


– More item base, so need optimized solution
–  Old user have more info but new user is not the case
–  Give new suggestions all the time
–  Discover all the products to user

Recommendation Approach


– Traditional collaborative filtering
– Suggestion based on similar customer
– Cluster model
– Suggestion based on similar customer
– Search based model
– Suggestion based on similar items
– Item to item collaborative filtering
– Suggestion based on similar items

Traditional collaborative filtering:


– Recommended predicted based on user similarity
– Users are grouped based on items/rating from there purchase/Viewed
– Similarity of users are determined by below formula,
– cos⁡(A,B)(A.B)/(|(|A|)|*||B||)
– select recommendations from the similar customers’ items using various methods as well, a common technique is to rank each item according to how many similar customers purchased it
TCF
Cluster model


– Group similar users as segments, Its a time taking process
– Require more processing time as we have more than million items/video
–  Not give fresh items all the time
–  Will not give appropriate related items as it grouped by users not by items

CM

Search based


– Suggesting items based on previously purchased item
– User has purchased item can get more suggestion than none for new users
– Suggestions are limited by purchased item
– No discover of new items in category

SB

Item to item collaborative filtering


– Items are grouped based on similarity (i.e,) mobile phone need case so suggest the items
– No user dependency on recommendation, so possible discovery of new items all the time
– Recommends highly correlated similar items, recommendation quality
is excellent
– Similarity between product is determined by,

For each item in product catalog, I1
For each customer C who purchased I1
For each item I2 purchased by customer C
Record that a customer purchased I1 and I2
For each item I2
Compute the similarity between I1 and I2

ITICF

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s