?7emsp;本文綜合整理了一些關(guān)于推薦算法的資料，資料來(lái)源注明在文章尾。

• Books：

1.《推薦系統(tǒng)實(shí)踐》項(xiàng)亮
??入門(mén)級(jí)教材咕宿，很薄，可以很快就看完蜡秽，把很多基礎(chǔ)而簡(jiǎn)單的問(wèn)題講的很詳細(xì)府阀。總體來(lái)說(shuō)芽突，此書(shū)性價(jià)比很高试浙，值得入手一本研讀
我買書(shū)喜歡上亞馬遜, 因?yàn)閬嗰R遜上很多都可以試讀，這本書(shū)亞馬遜就提供了試讀寞蚌，推薦大家先去試讀下田巴，再?zèng)Q定有沒(méi)有購(gòu)買價(jià)值。

2.《Recommender Systems Handbook》Paul B. Kantor
??有這本書(shū)就不用其它的了挟秤，很細(xì)很全固额，就是英文原版的有點(diǎn)小貴，真有志于做推薦系統(tǒng)的才去買吧煞聪，用到哪就翻書(shū)查。按人家的說(shuō)法逝慧，所有敢自稱handbook的書(shū)都是神書(shū)昔脯，沒(méi)看過(guò)這本書(shū)出去吹牛逼時(shí)你都不好意思說(shuō)自己是做推薦的。

3. Programming collective intelligence: building smart web 2.0 applications[M]
??寓教于樂(lè)的一本入門(mén)教材笛臣，附有可以直接動(dòng)手實(shí)踐的toy級(jí)別代碼

4. Jannach D, Zanker M, Felfernig A, et al. Recommender systems: an introduction[M]. Cambridge University Press, 2010
??可以認(rèn)為是2010年前推薦系統(tǒng)論文的綜述集合

5. Celma O. Music recommendation and discovery[M]. Springer, 2010
??主要內(nèi)容集中在音樂(lè)推薦云稚，領(lǐng)域非常專注于音樂(lè)推薦，包括選取的特征沈堡，評(píng)測(cè)時(shí)如何考慮音樂(lè)因素静陈。

6. Word sense disambiguation: Algorithms and applications[M]. Springer Science+ Business Media,2006
??如果涉及到關(guān)鍵詞推薦，或是文本推薦， 則可以查閱該書(shū)鲸拥。

• Conference & Journal:

• the ACM Conference on Recommender Systems,Recsys
  ?鏈接：https://recsys.acm.org/
• KDD
  ?鏈接：http://www.kdd.org/
• Special Interest Group on Information Retrieval,SIGIR
  ?鏈接：http://sigir.org/

• Papers：

綜述類:

1. 2002 - Hybrid Recommender Systems: Survey and Experiments
2. 2005 - Towards the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions拐格。最經(jīng)典的推薦算法綜述
3. 2009 - 個(gè)性化推薦系統(tǒng)的研究進(jìn)展.周濤等
4. 2011 - Collaborative Filtering Recommender Systems. JB Schafer 關(guān)于協(xié)同過(guò)濾最經(jīng)典的綜述
5. 2012 - 項(xiàng)亮的博士論文《動(dòng)態(tài)推薦系統(tǒng)關(guān)鍵技術(shù)研究》
6. 2012 - Recommender systems L Lü, M Medo, CH Yeung, YC Zhang, ZK Zhang, T Zhou Physics Reports 519 (1), 1-49 （https://arxiv.org/abs/1202.1112）
7. 2017-基于深度學(xué)習(xí)的推薦系統(tǒng)_黃立威

協(xié)同過(guò)濾：

1.matrix factorization techniques for recommender systems. Y Koren
2.Using collaborative filtering to weave an information Tapestry. David Goldberg （協(xié)同過(guò)濾第一次被提出）
3.Item-Based Collaborative Filtering Recommendation Algorithms. Badrul Sarwar , George Karypis, Joseph Konstan .etl
4.Application of Dimensionality Reduction in Recommender System -- A Case Study. Badrul M. Sarwar, George Karypis, Joseph A. Konstan etl
5.Probabilistic Memory-Based Collaborative Filtering. Kai Yu, Anton Schwaighofer, Volker Tresp, Xiaowei Xu,and Hans-Peter Kriegel
6.Recommendation systems:a probabilistic analysis. Ravi Kumar Prabhakar Raghavan.etl
7.Amazon.com recommendations: item-to-item collaborative filtering. Greg Linden, Brent Smith, and Jeremy York
8.Evaluation of Item-Based Top- N Recommendation Algorithms. George Karypis
9.Probabilistic Matrix Factorization. Ruslan Salakhutdinov
10.Tensor Decompositions,Alternating Least Squares and other Tales. Pierre Comon, Xavier Luciani, André De Almeida

基于內(nèi)容的推薦：

1.Content-Based Recommendation Systems. Michael J. Pazzani and Daniel Billsus

基于標(biāo)簽的推薦：

1.Tag-Aware Recommender Systems: A State-of-the-Art Survey. Zi-Ke Zhang(張子柯), Tao Zhou(周 濤), and Yi-Cheng Zhang(張翼成)

推薦評(píng)估指標(biāo)：

1、推薦系統(tǒng)評(píng)價(jià)指標(biāo)綜述. 朱郁筱刑赶，呂琳媛
2捏浊、Accurate is not always good：How Accuacy Metrics have hurt Recommender Systems
3、Evaluating Recommendation Systems. Guy Shani and Asela Gunawardana
4撞叨、Evaluating Collaborative Filtering Recommender Systems. JL Herlocker

推薦多樣性和新穎性：

1. Improving recommendation lists through topic diversification. Cai-Nicolas Ziegler
   Sean M. McNee, Joseph A.Konstan,Georg Lausen
2. Fusion-based Recommender System for Improving Serendipity
3. Maximizing Aggregate Recommendation Diversity：A Graph-Theoretic Approach
4. The Oblivion Problem：Exploiting forgotten items to improve Recommendation diversity
5. A Framework for Recommending Collections
6. Improving Recommendation Diversity. Keith Bradley and Barry Smyth

推薦系統(tǒng)中的隱私性保護(hù)：

1.Collaborative Filtering with Privacy. John Canny
2.Do You Trust Your Recommendations? An Exploration Of Security and Privacy Issues in Recommender Systems. Shyong K “Tony” Lam, Dan Frankowski, and John Ried.
3.Privacy-Enhanced Personalization. Alfred Kobsa.etl
4.Differentially Private Recommender Systems：Building Privacy into the
Netflix Prize Contenders. Frank McSherry and Ilya Mironov Microsoft Research,
Silicon Valley Campus
5.When being Weak is Brave: Privacy Issues in Recommender Systems. Naren Ramakrishnan, Benjamin J. Keller,and Batul J. Mirza

推薦冷啟動(dòng)問(wèn)題：

1.Tied Boltzmann Machines for Cold Start Recommendations. Asela Gunawardana.etl
2.Pairwise Preference Regression for Cold-start Recommendation. Seung-Taek Park, Wei Chu
3.Addressing Cold-Start Problem in Recommendation Systems. Xuan Nhat Lam.etl
4.Methods and Metrics for Cold-Start Recommendations. Andrew I. Schein, Alexandrin P opescul, Lyle H. U ngar

bandit(老虎機(jī)算法,可緩解冷啟動(dòng)問(wèn)題):

1.Bandits and Recommender Systems. Jeremie Mary, Romaric Gaudel, Philippe Preux
2.Multi-Armed Bandit Algorithms and Empirical Evaluation
基于社交網(wǎng)絡(luò)的推薦：

1. Social Recommender Systems. Ido Guy and David Carmel
2. A Social Networ k-Based Recommender System(SNRS). Jianming He and Wesley W. Chu
3. Measurement and Analysis of Online Social Networks.
4. Referral Web：combining social networks and collaborative filtering
   基于知識(shí)的推薦：
   1.Knowledge-based recommender systems. Robin Burke
   2.Case-Based Recommendation. Barry Smyth
   3.Constraint-based Recommender Systems: Technologies and Research Issues. A. Felfernig. R. Burke
   其他：
   Trust-aware Recommender Systems. Paolo Massa and Paolo Avesani
   推薦幾篇對(duì)工業(yè)界比較有影響的論文吧：
5. The Wisdom of The Few 豆瓣阿穩(wěn)在介紹豆瓣猜的時(shí)候極力推薦過(guò)這篇論文金踪，豆瓣猜也充分應(yīng)用了這篇論文中提出的算法；
6. Restricted Boltzmann Machines for Collaborative Filtering 目前Netflix使用的要推薦算法之一牵敷；
7. Factorization Meets the Neighborhood: a Multifaceted Collaborative Filtering Model 這個(gè)無(wú)需強(qiáng)調(diào)重要性胡岔，LFM幾乎應(yīng)用到了每一個(gè)商業(yè)推薦系統(tǒng)中；
8. Collaborative Filtering with Temporal Dynamics 加入時(shí)間因素的SVD++模型枷餐，曾在Netflix Prize中大放溢彩的算法模型靶瘸；
9. Context-Aware Recommender Systems 基于上下文的推薦模型，現(xiàn)在不論是工業(yè)界還是學(xué)術(shù)界都非臣馓裕火的一個(gè)topic奕锌；
10. Toward the next generation of recommender systems 對(duì)下一代推薦系統(tǒng)的一個(gè)綜述；
11. Item-Based Collaborative Filtering Recommendation Algorithms 基于物品的協(xié)同過(guò)濾村生，Amazon等電商網(wǎng)站的主力模型算法之一惊暴；
12. Information Seeking-Convergence of Search, Recommendations and Advertising 搜索、推薦和廣告的大融合也是未來(lái)推薦系統(tǒng)的發(fā)展趨勢(shì)之一趁桃；
13. Ad Click Prediction: a View from the Trenches 可以對(duì)推薦結(jié)果做CTR預(yù)測(cè)排序辽话；
14. Performance of Recommender Algorithm on top-n Recommendation Task TopN預(yù)測(cè)的一個(gè)綜合評(píng)測(cè)，TopN現(xiàn)在是推薦系統(tǒng)的主流話題卫病，可以全部實(shí)現(xiàn)這篇文章中提到的算法大概對(duì)TopN有個(gè)體會(huì)油啤；
15. http://dsec.pku.edu.cn/~jinlong/publication/wjlthesis.pdf 北大一博士對(duì)Netflix Prize算法的研究做的畢業(yè)論文，這篇論文本身對(duì)業(yè)界影響不大蟀苛，但是Netflix Prize中運(yùn)用到的算法極大地推動(dòng)了推薦系統(tǒng)的發(fā)展益咬；
    通過(guò)這些論文可以對(duì)推薦系統(tǒng)有個(gè)總體上的全面認(rèn)識(shí)，并且能夠了解一些推薦系統(tǒng)的發(fā)展趨勢(shì)帜平。剩下的就是多實(shí)踐了幽告。

推薦兩篇必看（最好能自己實(shí)現(xiàn)）論文， 其他的論文其實(shí)都是在這基礎(chǔ)上build起來(lái)的裆甩。
http://Amazon.com Recommendations Item-to-Item Collaborative Filtering
http://www.cin.ufpe.br/~idal/rs/Amazon-Recommendations.pdf

MATRIX FACTORIZATION TECHNIQUES FOR RECOMMENDER SYSTEMS
https://datajobs.com/data-science-repo/Recommender-Systems-%5BNetflix%5D.pdf
推薦兩篇必看（最好能自己實(shí)現(xiàn)）論文冗锁， 其他的論文其實(shí)都是在這基礎(chǔ)上build起來(lái)的。
http://Amazon.com Recommendations Item-to-Item Collaborative Filtering
http://www.cin.ufpe.br/~idal/rs/Amazon-Recommendations.pdf

MATRIX FACTORIZATION TECHNIQUES FOR RECOMMENDER SYSTEMS
https://datajobs.com/data-science-repo/Recommender-Systems-%5BNetflix%5D.pdf

• Articles：

基于用戶投票的排名算法（一）：Delicious和Hacker News http://www.ruanyifeng.com/blog/2012/02/ranking_algorithm_hacker_news.html

youtube的推薦算法經(jīng)歷過(guò)好幾次大的改動(dòng)嗤栓，都有論文發(fā)表的：https://www.zhihu.com/question/20829671

Netflix推薦算法冻河，讓每個(gè)人看到不一樣的電影海報(bào)：https://juejin.im/post/5a2e71e351882575d42f5651

淘寶網(wǎng)的推薦算法：https://www.zhihu.com/question/29108284

• Website：

• leetcode箍邮，最近很火的算法網(wǎng)站：https://leetcode.com/
  
  

• Blog:

• 一些數(shù)據(jù)挖掘與分析文章的翻譯(https://github.com/ictar/python-doc/tree/master/Science%20and%20Data%20Analysis)
• Recommending music on Spotify with deep learning
• BreezeDeus
• 餓了么推薦系統(tǒng)：從0到1 - 極客頭條 - CSDN.NET
  https://github.com/ictar/python-doc/tree/master/Science%20and%20Data%20Analysis

• Recommendation Engines Seminar Paper, Thomas Hess, 2009: 推薦引擎的總結(jié)性文章，Thomas 給出推薦引擎的模型叨叙，各種推薦機(jī)制的工作原理锭弊，并分析了推薦引擎面臨的眾多問(wèn)題。
• Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions, Adomavicius, G.; Tuzhilin, A., 2005：2005 年的論文摔敛，對(duì)現(xiàn)今流行的推薦技術(shù)進(jìn)行總結(jié)廷蓉，深入具體的實(shí)現(xiàn)技術(shù)方法，同時(shí)也提出了對(duì)下一代推薦引擎的展望马昙。
• A Taxonomy of RecommenderAgents on the Internet, Montaner, M.; Lopez, B.; de la Rosa, J. L., 2003, 對(duì)互聯(lián)網(wǎng)上推薦引擎進(jìn)行總結(jié)桃犬，給出的不同推薦方法的分類和特點(diǎn)，幫助讀者對(duì)推薦引擎有全面的認(rèn)識(shí)行楞。
• Amazon: www.amazon.com：推薦技術(shù)的先驅(qū)攒暇，Amazon 在 B2C 領(lǐng)域的推薦技術(shù)值得大家參考。
• 豆瓣：www.douban.com：作為國(guó)內(nèi)社交網(wǎng)絡(luò)的先驅(qū)子房，豆瓣在推薦技術(shù)上也處于領(lǐng)先的位置形用，同時(shí)對(duì)于不同內(nèi)容的推薦策略有深入的研究。
• 個(gè)性化推薦技術(shù)漫談：對(duì)個(gè)性化推薦技術(shù)的基本原理進(jìn)行簡(jiǎn)要介紹证杭，提出了作者對(duì)優(yōu)秀的個(gè)性化推薦的多角度認(rèn)識(shí)
• Google Recommender System Group：推薦系統(tǒng)的 Google 討論組田度，有很多關(guān)于推薦引擎的有趣討論
• Recommender System Algorithms：關(guān)于推薦引擎算法的資源
• Design of Recommender System：關(guān)于推薦引擎的設(shè)計(jì)方法的介紹
• How to build a recommender system：這個(gè)演示給出了如何構(gòu)建一個(gè)推薦引擎，并結(jié)合例子詳細(xì)介紹了基于協(xié)同過(guò)濾的推薦策略解愤。
  
  

• Lecture:

• 【直播回顧】21天搭建推薦系統(tǒng)：實(shí)現(xiàn)“千人千面”個(gè)性化推薦（含視頻）-博客-云棲社區(qū)-阿里云
  
  

• Code：

• 不到100行代碼實(shí)現(xiàn)一個(gè)簡(jiǎn)單的推薦系統(tǒng)

• 利用Python實(shí)現(xiàn)基于協(xié)同過(guò)濾算法的影片推薦

• 使用Spark MLlib給豆瓣用戶推薦電影

• Python中一個(gè)簡(jiǎn)單的基于內(nèi)容的推薦引擎

• https://github.com/MashiMaroLjc/ML-and-DM-in-action
  https://github.com/laozhaokun/movie_recommender

• 利用Python實(shí)現(xiàn)基于協(xié)同過(guò)濾算法的影片推薦

• 推薦算法庫(kù)：

SvdFeature
LibFM
Mahout
MLib

reference:

• https://www.zhihu.com/question/21251105/answer/21094996
• https://www.zhihu.com/question/25566638/answer/337012082
• https://www.zhihu.com/question/25566638/answer/37455091