A Study On Social-based Feed Recommendation Service For S Company

In recent years,the booming development of information technology and the Internet industry has brought human society into the era of big data.It is important to maximize the value of the data and let users efficiently find content that matches their interest preferences from the overload information,which improving product differentiation and competitiveness.From this point of view,recommendation system is one of the best solutions for this scenario.This article takes S company's core product “MP App” as an example.Firstly,it introduces the basic situation and operation management of the product.With conducting a comprehensive and systematic analysis of the internal and external environment in which the product is located,further understanding of the product internal logic and user demand is clarified;secondly,the product operation status is deeply analyzed,and the problems existing in the current product operation in the work process and method are deeply explored,the strategic significance of the feeds recommendation system for product operation management and company development is investigated.Furthermore,it discusses the construction process of the feeds recommendation system,combined with the characteristics of product data model and distribution.In order to address the problem of matrix sparsity and utilize the social network properties,a SOMF model is introduced,which is original from basic matrix factorization form.On the other hand,the optimization strategy of the model is optimized,the influence of different hyper-parameters on the model results is discussed,and the methods of online testing and effect evaluation of the recommended system are summarized.Finally,in combination with the current actual work,suggestions and countermeasures for strengthening the data infrastructure and using data to empower the business are proposed.