Recommendtion System Introduction

最新推荐文章于 2022-09-13 01:31:35 发布

原创最新推荐文章于 2022-09-13 01:31:35 发布 · 197 阅读

0 ·

CC 4.0 BY-SA版权

Mahout 专栏收录该内容

25 篇文章

订阅专栏

本文介绍了推荐系统的两种主要算法：基于用户的协同过滤和基于内容的推荐。基于用户的协同过滤依赖于用户与物品之间的关系来生成推荐，而不需要了解物品本身的特性。基于内容的推荐则依据物品属性进行推荐，这种方法虽然有效但具有领域局限性。

collaborative filtering

producing recommendations based on, and only based on, knowledge of users’ rela-tionships to items. These techniques require no knowledge of the properties of the items themselves.

In fact, these are the two broadest categories of recommender engine algorithms: user-based and item-based recommenders.

content-based recommendation

based on the attributes of items. There’s nothing wrong with content-based techniques; on the contrary, they can work quite well. They’re necessarily domain-specific approaches, and they’d be hard to meaningfully codify into a framework. To build an effective content-based book rec-ommender, one would have to decide which attributes of a book—page count, author,publisher, color, font—are meaningful, and to what degree. None of this knowledgetranslates into any other domain; recommending books this way doesn’t help in the area of recommending pizza toppings.

Reference

<<Mahout In Action>>