View
464
Download
2
Category
Tags:
Preview:
DESCRIPTION
Video available here http://www.youtube.com/watch?v=1jHxGCl8RXc Recommender systems research is often based on comparisons of predictive accuracy: the better the evaluation scores, the better the recommender. However, it is difficult to compare results from different recommender systems due to the many options in design and implementation of an evaluation strategy. Additionally, algorithmic implementations can diverge from the standard formulation due to manual tuning and modifications that work better in some situations. In this work we compare common recommendation algorithms as implemented in three popular recommendation frameworks. To provide a fair comparison, we have complete control of the evaluation dimensions being benchmarked: dataset, data splitting, evaluation strategies, and metrics. We also include results using the internal evaluation mechanisms of these frameworks. Our analysis points to large differences in recommendation accuracy across frameworks and strategies, i.e. the same baselines may perform orders of magnitude better or worse across frameworks. Our results show the necessity of clear guidelines when reporting evaluation of recommender systems to ensure reproducibility and comparison of results.
Citation preview
Comparative Recommender System Evaluation: Benchmarking Recommendation Frameworks
Alan Said@alansaid
TU Delft
Alejandro Bellogín@abellogin
Universidad Autónoma de Madrid
ACM RecSys 2014Foster City, CA, USA
2
A RecSys paper outline– We have a new model – it’s
great– We used %DATASET% 100k
to evaluate it– It’s 10% better than our
baseline– It’s 12% better than
[Authors, 2010]
3
Benchmarking
4
LibRecmrec
Python-recsys
5
What are the differences?Some things just work differently• Data splitting• Algorithm design (implementation)• Algorithm optimization• Parameter values• Evaluation• Relevance/ranking• Software architecture• etc
Different design choices!!
How do these choices affect evaluation results?
6
Evaluate evaluation• Comparison of frameworks• Comparison of implementation• Comparison of results• Objective benchmarking
7
Algorithmic ImplementationFramework Class Similarity Item-based
LensKit ItemItemScorer CosineVectorSimilarity, PearsonCorrelation
Mahout GenericItemBasedRecommender UncenteredCosineSimilarity, PearsonCorrelationSimilarityMyMediaLite ItemKNN Cosine, Pearson User-based Parameters
LensKit UserUserItemScorerCosineVectorSimilarity, PearsonCorrelation
SimpleNeighborhoodFinder, NeighborhoodSize
Mahout GenericUserBasedRecommenderUncenteredCosineSimilarity, PearsonCorrelationSimilarity
NearestNUserNeighborhood, neighborhoodsize
MyMediaLite UserKNN Cosine, Pearson neighborhoodsize Matrix Factorization
LensKit FunkSVDItemScorer IterationsCountStoppingCondition, factors, iterations
Mahout SVDRecommender FunkSVDFactorizer, factors, iterationsMyMediaLite SVDPlusPlus factors, iterations
8
There’s more than algorithms thoughThere’s the data, evaluation, and more
Data splits• 80-20 Cross-validation• Random Cross-validation• User-based cross validation• Per-user splits• Per-item splits• Etc.
Evaluation• Metrics• Relevance• Strategies
9
Real world examples
Movielens 1M[Cremonesi et al, 2010]
Movielens 1M[Yin et al, 2012]
10
Evaluation
Dataset
Training / Test Framework Evaluation
Results
Algorithm
11
Internal Evaluation
Dataset
Training / Test Framework Evaluation
Results
Algorithm
12
Internal Evaluation ResultsAlgorithm Framework nDCG
IB Cosine Mahout 0,00041478
IB Cosine Lenskit 0,94219205
IB Pearson Mahout 0,00516923
IB Pearson Lenskit 0,92454613
SVD50 Mahout 0,10542729
SVD50 Lenskit 0,94346409
UB Cosine Mahout 0,16929545
UB Cosine Lenskit 0,94841356
UB Pearson Mahout 0,16929545
UB Pearson Lenskit 0,94841356
Algorithm Framework RMSEIB Cosine Lenskit 1,01390931
IB Cosine MyMediaLite 0,92476162
IB Pearson Lenskit 1,05018614
IB Pearson MyMediaLite 0,92933246
SVD50 Lenskit 1,01209290
SVD50 MyMediaLite 0,93074012
UB Cosine Lenskit 1,02545490
UB Cosine MyMediaLite 0,93419026
13
We need a fair and common evaluation protocol!
14
Reproducible evaluation - BenchmarkingControl all parts of the process
- Data Splitting strategy- Recommendation (black box)- Candidate items generation (what
items to test)- Evaluation
Select strategy• By time• Cross validation• Random• Ratio
Select framework• Apache Mahout• LensKit• MyMediaLite
Select algorithm• Tune settings
Recommend
Define strategy• What is the
ground truth• What users to
evaluate• What items to
evaluate
Select error metrics• RMSE, MAE
Select ranking metrics• nDCG,
Precision/Recall, MAP
RecommendSplitCandidate
itemsEvaluate
http://rival.recommenders.net
15
Controlled Evaluation
Dataset
Training / Test Framework Evaluation
Results
Algorithm
16
AN OBJECTIVE BENCHMARK
Lenskit vs. Mahout vs. MyMediaLiteMovielens 100k (additional datasets in the paper)
17
The FrameworksAM: Apache MahoutLK: LenskitMML: MyMediaLite
The Candidate ItemsRPN: Relevant + N [Koren, KDD 2008]TI: TrainItemsUT: UserTest
Split Pointgl: Globalpu: Per-user
Split Strategycv: 5-fold cross-validationrt: 80-20 random ratio
Algorithms nDCG@10
18
User Coverage
19
Catalog Coverage
20
Time
21
Good accuracy?
22
Yes, at the cost of coverage
23
What’s the best result?
24
Difficult to say … depends on what you’re evaluating!!
25
In conclusion• Design choices matter!
– Some more than others• Evaluation needs to be documented• Cross-framework comparison is not easy
– You need to have control!
26
What have we learnt?
27
How did we do this?RiVal – an evaluation toolkit for RecSys• http://rival.recommenders.net• http://github.com/recommenders/rival• RiVal demo later today• On Maven central!
• RiVal was also used for this year’s RecSys Challenge
– www.recsyschallenge.com
28
QUESTIONS?Thanks!
Special thanks: • Zeno Gantner• Michael Ekstrand
29
• https://www.flickr.com/photos/13698839@N00/3001363490/in/photostream/• http://rick--hunter.deviantart.com/art/Unfair-scale-1-149667590
Recommended