摘要随着web2.0的发展成熟,人们逐渐进入信息过载的时代。在信息消费者面对大量信息无所适从的同时,信息生产者也开始烦恼如何吸引广大消费者的目光。为解决这一矛盾,推荐系统应运而生。它作为文系两者关系的一个桥梁,既能帮助消费者寻到符合心中所想的信息,又能让生产者得到更多人的关注,实现了双赢。本文的研究工作的主要集中在以下三个方面:
一方面,本文归纳总结了现有个性化推荐系统技术与方法,收集、整理了推荐系统的常见数据集、国内外相关竞赛。另一方面,本文重点关注了协同过滤的推荐算法。包括基于用户的协同过滤和基于项目的协同过滤,并进行实验仿真。最后,实现了一个简单的基于Apache Mahout的电影推荐引擎,包括User CF、Item CF以及Slope One算法的推荐列表。26999
关键词 个性化推荐系统 协同过滤 Apache Mahout 毕业论文设计说明书外文摘要
Title Merchandise personalized recommendation system for electric business platform
Abstract
With the development of web 2.0, people gradually enter into the era of information overload. When the information consumers were lost in large amounts of information, information producers also began to worry about how to attract the attention of consumers. To solve this contradiction, the recommendation system arises at the historic moment. As a bridge to maintain the relationship between them, it both can help consumers find the information they want, but also allows producers to get more attention, which achieves a win-win situation.
In this paper, the research work mainly focused on the following three aspects:
On the one hand, this paper summarizes the existing personalized recommendation system technologies and methods,collects and sorts out common datasets, domestic and foreign competitions.
On the other hand, this article focuses on the collaborative filtering recommendation algorithm. Including Item-based collaborative filtering and User-based collaborative filtering, and the simulation experiment.
Finally, I have achieved a simple film-based Apache Mahout recommendation engine, including the recommendation list of User-based CF, Item-based CF and Slope One algorithm.
Keywords personalized recommendation system collaborative filtering Apache Mahout
目 次
1 绪论 1
1.1 课题研究背景及意义 1
1.2 研究现状 2
1.3 论文的研究内容与组织 2
2 推荐系统综述及相关技术 4
2.1 推荐系统组成 4
2.2 推荐系统分类 4
2.2.1 基于内容的推荐系统 5
2.2.2 基于协同过滤的推荐系统 5
2.2.3 基于关联规则的推荐系统 5
2.2.4 基于知识的推荐系统 5
2.3 推荐系统评价方法 5
2.3.1 线下实验(Offline Experiment) 6
2.3.2 用户调查问卷(User Survey) 6
2.3.3 线上实验(On-line Experiment) 6
2.4 推荐系统评测指标 6
2.4.1 准确率(Precision) 6
2.4.2 召回率(Recall) 6
2.4.3 覆盖率(Coverage) 7
2.5 推荐系统常用数据集及国内外相关竞赛 7
2.6 Apache Mahout 介绍 7
3 协同过滤推荐算法的研究 10
3.1 协同过滤核心 10