基于 SVD 的協(xié)同過(guò)濾電影推薦算法

打開(kāi)文本圖片集

摘要：信息過(guò)載的問(wèn)題愈發(fā)嚴(yán)重在大數(shù)據(jù)時(shí)代針對(duì)不同用戶(hù)提高電影推薦系統(tǒng)的推薦性能一直存在巨大的挑戰(zhàn)為了有效地解決信息過(guò)載和用戶(hù)體驗(yàn)滿意度低的問(wèn)題需要選擇合適的個(gè)性化推薦算法文章概述了主流的機(jī)器學(xué)習(xí)推薦算法并通過(guò)實(shí)驗(yàn)比較分析了各算法的優(yōu)缺點(diǎn)針對(duì)推薦算法普遍存在的冷啟動(dòng)和數(shù)據(jù)稀疏性問(wèn)題提出了相應(yīng)的解決方案

關(guān)鍵詞：電影推薦;協(xié)同過(guò)濾（CF）;冷啟動(dòng);奇異值分解（SVD）

中圖法分類(lèi)號(hào)：TP391文獻(xiàn)標(biāo)識(shí)碼：A

Collaborative filtering movie recommendationalgorithm based on SVD

SONG Longsheng'，WANG Jialel，NI Shengqiao1.2

（1.College of Information Science and Technology，Tibet University，Lhasa 850000，China：

2.College of Computer Science，Sichuan University，Chengdu 610065，China）

Abstract：Nowadays， the problem of information overload is becoming more and more serious. In the sea of big data， it has been a huge challenge to improve the recommendation performance of moie recommendation system for different users. In order to effectively solve the problems of information overload and low user experience satisfaction， it is necessary to choose an appropriate personalized recommendation algorithm. In this paper， the mainstream machine learning recommendation algorithms are summarized， and the advantages and disadvantages of each algorithm are compared and analyzed through experiments. Finally， corresponding solutions are proposed to solve the common problems of cold start and data sparsity in recommendation algorithms.

Key words： film recommendation， collaborative filtering（CF）， cold start， singular value decomposition（SVD）

1  引言

近年來(lái)，隨著互聯(lián)網(wǎng)的飛速發(fā)展，影視產(chǎn)業(yè)的數(shù)量和種類(lèi)激增且發(fā)展迅猛，“電影過(guò)載”的問(wèn)題愈發(fā)嚴(yán)重，導(dǎo)致用戶(hù)尋找自己喜歡的電影需要浪費(fèi)很多時(shí)間。（剩余3815字）