ABSTRACT
In this paper we introduce RecLab, a system designed to enable developers to build and test recommendation algorithms for eCommerce websites. RecLab supports a variety of context and feedback that recommenders can take advantage of to improve the quality of their recommendations. RecLab is unique in that recommenders built on top of RecLab APIs can run in the RichRelevance cloud environment in addition to working with offline data sets. This environment provides on-site recommendations to some of the largest eCommerce sites in existence. By running in the cloud, we are able to avoid the pitfalls that have historically made it difficult for researchers to work with real data and live traffic. We bring code to data, rather than bringing sensitive merchant data to code running outside the cloud. Peer-reviewed recommenders built with RecLab will be chosen to run in the cloud on live data, given their authors unprecedented feedback into their performance in situ.
- Marshall, Matt. Aggregate Knowledge raises $5M from Kleiner, on a roll. Venture Beat http://venturebeat.com/2006/12/10/aggregate-knowledge-raises-5m-from-kleiner-on-a-roll/(Dec. 10, 2006).Google Scholar
- Herlocker, Jonathan L., Konstan, Joseph A., Terveen, Loren G., and Riedl, John T. Evaluating collaborative filtering recommender systems. ACM Trans. Inf. Syst., 22, 1 (2004), 5--53. Google ScholarDigital Library
- Gunawardana, Asela and Shani, Guy. A Survey of Accuracy Evaluation Metrics of Recommendation Tasks. J. Machine Learning Research, 10 (2009), 2935--2962. Google ScholarDigital Library
- Shani, Guy and Gunawardana, Asela. Evaluating Recommendation Systems. In Ricci, Francesco et al., eds., Recommender Systems Handbook. Springer, 2011.Google ScholarCross Ref
- Pu, Pearl and Chen, Li and Kumar, Pratyush. Evaluating product search and recommender systems for E-commerce environments. Electronic Commerce Research, 8, 1--2 (2008), 1--27. Google ScholarDigital Library
- Schafer, J. Ben, Konstan, Joseph A., and Riedl, John. E-Commerce Recommendation Applications. Data Mining and Knowledge Discovery, 5, 1--2 (2001), 115--153. Google ScholarDigital Library
- Kent, Allen, Berry, Madeline M., Luehrs, Fred U., and Perry, J. W. Machine literature searching VIII. Operational criteria for designing information retrieval systems. American Documentation, 6, 2 (1955), 93--101.Google ScholarCross Ref
- Ziegler, Cai-Nicolas and McNee, Sean M. and Konstan, Joseph A. and Lausen, Georg. Improving recommendation lists through topic diversification. In Proc. 14th int'l conf. on World Wide Web (2005), ACM, 22--32. Google ScholarDigital Library
- Akiyama, Takayuki, Obara, Kiyohiro, and Tanizaki, Masaaki. Proposal and Evaluation of Serendipitous Recommendation Method Using General Unexpectedness. In Proc. Workshop on the Practical Use of Recommender Systems, Algorithms and Technologies (2010), CEUR, 3--10.Google Scholar
- Zheng, Hua, Wang, Dong, Zhang, Qi, Li, Hang, and Yang, Tinghao. Do clicks measure recommendation relevancy?: an empirical user study. In Proc. of the Fourth ACM Conf. on Recommender Systems (2010), ACM, 249--252. Google ScholarDigital Library
- Jones, Nicolas and Pu, Pearl. User Technology Adoption Issues in Recommender Systems. In Proc. Networking and Electronic Commerce Research Conference (NAEC) (2007), 379--394.Google Scholar
- Pu, Pearl and Chen, Li. A User-Centric Evaluation Framework of Recommender Systems. In Proc. ACM RecSys 2010 Workshop on User-Centric Evaluation of Recommender Systems and Their Interfaces (UCERSTI) (2010). Google ScholarDigital Library
- Adomavicius, Gediminas and Tuzhilin, Alexander. Context-Aware Recommender Systems. In Ricci, Francesco et al., eds., Recommender Systems Handbook. Springer, 2011.Google Scholar
- Anand, Sarabjot Singh and Mobasher, Bamshad. Contextual Recommendation. In Berendt, B. et al., eds., From Web to Social Web: Discovering and Deploying User and Content Profiles. Springer, 2007. Google ScholarDigital Library
- Jäschke, Robert, Eisterlehner, Folke, Hotho, Andreas, and Stumme, Gerd. Testing and Evaluating Tag Recommenders in a Live System. In Workshop on Knowledge Discovery, Data Mining, and Machine Learning (2009), 44--51.Google ScholarDigital Library
- Naughton, Robin and Lin, Xia. Recommender Systems: Investigating the Impact of Recommendations on User Choices and Behaviors. In Proc. ACM RecSys 2010 Wkshp on User-Centric Evaluation of Recommender Systems and Their Interfaces (UCERSTI) (2010).Google Scholar
- Castagnos, Sylvain, Jones, Nicolas, and Pu, Pearl. Eye-tracking product recommenders' usage. In Proc. Fourth ACM Conf. on Recommender Systems (2010), ACM, 28--36. Google ScholarDigital Library
- Ricci, Francesco et al., eds. Recommender Systems Handbook. Springer, 2011. Google ScholarDigital Library
- Dahlen, Brent J., Konstan, Joseph, Herlocker, Jon, Good, Nathaniel, Borchers, Al, and Riedl, John. Jump-Starting MovieLens: User Benefits of Starting a Collaborative Filtering System with "Dead Date". Technical Report 98--017, University Of Minnesota, 1998.Google Scholar
- Herlocker, Jonathan L., Konstan, Joseph A., Borchers, Al, and Riedl, John. An algorithmic framework for performing collaborative filtering. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval (SIGIR) (1999), ACM, 230--237. Google ScholarDigital Library
- Lohr, Steve. Netflix Cancels Contest After Concerns Are Raised About Privacy. New York Times (March 12, 2010).Google Scholar
- Zeller, Tom. AOL Technology Chief Quits After Data Release. New York Times (2006).Google Scholar
- Selinger, David, Kohn, Tyler, Liu, Wendy, and Burk, Scott. Resolving the Inherent Conflicts in Academic-Commercial Data Mining Collaboration. In Workshop on Data Mining for Business Applications (2006), ACM, 81--84.Google Scholar
- Linden, Greg. Slides from my talk at Stanford. Geeking with Greg http://glinden.blogspot.com/2006/12/slides-from-my-talk-at-stanford.html (December 4, 2006).Google Scholar
- Farber, Dan. Between the Lines. Google's Marissa Mayer: Speed wins http://www.zdnet.com/blog/btl/googles-marissa-mayer-speed-wins/3925 (November 9, 2006).Google Scholar
- Apache License, Version 2.0 http://www.apache.org/licenses/LICENSE-2.0.html (2004).Google Scholar
- Dean, Jeffrey and Ghemawat, Sanjay. MapReduce: Simplified Data Processing on Large Clusters. In 6th Symposium on Operating System Design and Implementation (2004), USENIX, 137--150. Google ScholarDigital Library
- Zhao, Zhi-Dan and Shang, Ming-sheng. User-Based Collaborative-Filtering Recommendation Algorithms on Hadoop. In Proc. Third Intl. Conf. on Knowledge Discovery and Data Mining (2010), 478--481. Google ScholarDigital Library
- Liu, Chao, Yang, Hung-chih, Fan, Jinliang, He, Li-Wei, and Wang, Yi-Min. Distributed nonnegative matrix factorization for web-scale dyadic data analysis on mapreduce. In Proc. Intl. conf. on WWW (2010), 681--690. Google ScholarDigital Library
- snakeyaml http://code.google.com/p/snakeyaml/.Google Scholar
- Subversion http://subversion.tigris.org/.Google Scholar
- Apache Maven Project http://maven.apache.org/.Google Scholar
- Sarwar, Badrul, Karypis, George, Konstan, Joseph, and Reidl, John. Item-based collaborative filtering recommendation algorithms. In Proc. 10th intl. conf. on WWW (2001), 285--295. Google ScholarDigital Library
- Hadoop http://hadoop.apache.org/.Google Scholar
- Hadoop MapReduce http://hadoop.apache.org/mapreduce/.Google Scholar
- The Official YAML Website http://www.yaml.org/.Google Scholar
- Lemire, Daniel and Maclachlan, Anna. Slope One Predictors for Online Rating-Based Collaborative Filtering. In Proc. SIAM Data Mining (SDM) (2005), SIAM.Google Scholar
Index Terms
- RecLab: a system for eCommerce recommender research with real data, context and feedback
Recommendations
User-Oriented Context Suggestion
UMAP '16: Proceedings of the 2016 Conference on User Modeling Adaptation and PersonalizationRecommender systems have been used in many domains to assist users' decision making by providing item recommendations and thereby reducing information overload. Context-aware recommender systems go further, incorporating the variability of users' ...
Context suggestion: empirical evaluations vs user studies
WI '17: Proceedings of the International Conference on Web IntelligenceRecommender System has been successfully applied to assist user's decision making by providing a list of recommended items. Context-aware recommender system additionally incorporates contexts (such as time and location) into the system to improve the ...
Experimental evaluation of context-dependent collaborative filtering using item splitting
Collaborative Filtering (CF) computes recommendations by leveraging a historical data set of users' ratings for items. CF assumes that the users' recorded ratings can help in predicting their future ratings. This has been validated extensively, but in ...
Comments