From Text Mining to Recommendation Systems

Recommendation systems are the crucial component of e-commerce (suggest to buyers articles that could interest them) and online advertisement (suggest to users the right contents, matching their preferences). With the rise of many web services (Youtube, Amazon, Netflix ...) ,
recommender systems become unavoidable and unseparable in our daily life.
From the technical point of view, input data to recommendation algorithms are the utility matrices, which are similar to the bag of word representation of documents. We would like to recall some recommendation algorithms based on text mining techniques including LSA(Latent Semantic Analysis) and LDA (Latent Dirichlet Allocation).