Developer will receive an SQL dump of a table with two column, a userid and an object id. The job is to write four efficient algorithms.
recommended_object(int objectid) : returns an array of similar objects and a relevance score per.
recommended_person(int personid) : returns an array of similar people and a relevance score per.
compare_object(int objectid1, int objectid1) : returns a relevance score.
compare_people(int personid1, int personid2) : returns a relevance score based on the user's interests
To bid, please provide two to three sentences on how this is computationally accomplishable.
5 freelance font une offre moyenne de $132 pour ce travail
Dear sir, I can use Jaccard similarity measurement to solve it based on the SQL dump. The four algorithms is based on the Jaccard similarity to get the relevant objectIDs or UserIDs, so it is computational achievabl Plus
Hi, I would use permutations and probability. Though, what worries my is the practical viability of this..., what is the amount of data we are dealing with, should I worry about the processing time?
I am an experienced developer, and I'd be very interested in taking on this project. I am a detailed oriented, logical and creative thinker. Please see response in your Mailbox.
Will use Bayesian network to learn the user's and item's similarity. for a given user the we can sort item by probability p(a=1 | b) where b is all purchased items.. Had used this methods in music recommendation. ht Plus