基于word2vec的推荐系统设计

本文首先介绍了推荐系统出现的背景:信息过载而个人需求不明确;接着介绍了推荐系统背后的基本原理:基于相似度的推荐;然后介绍了相似度计算的几种方法,期中最为常用的是余弦相似度,然后简要介绍了Word2vec这一计算余弦相似度最常用的工具,以及其涉及到的符号向量化表示,概率语言模型,训练模型,作为抛砖引玉的例子,最后简单介绍了2个基于word2vec工具的推荐系统如何设计。

This paper first introduces the background of recommendation system:information overload and personal needs unclear;Then it introduces the basic principles behind the recommendation system:recommendation based on similarity;And then introduces the several methods of similarity calculation,is the most commonly used in the mid-term cosine similarity,and then briefly introduces the Word2vec this calculate the cosine similarity of the most commonly used tools,as well as its symbol to quantify involved,probability language model,training model,as examples,the author finally introduces the two recommendation system based on Word2vec tool design.