Difference between revisions of "User:Zeno Gantner"
Jump to navigation
Jump to search
Zeno Gantner (talk | contribs) |
Zeno Gantner (talk | contribs) |
||
(46 intermediate revisions by the same user not shown) | |||
Line 2: | Line 2: | ||
Primary developer of the [[MyMediaLite]] recommender system library. | Primary developer of the [[MyMediaLite]] recommender system library. | ||
Co-organizer of the [[Recommender Stammtisch]] in Berlin. | Co-organizer of the [[Recommender Stammtisch]] in Berlin. | ||
− | |||
− | |||
− | |||
− | |||
[http://www.ismll.uni-hildesheim.de/personen/gantner_en.html homepage], [https://scholar.google.com/citations?user=AhVYsaoAAAAJ Google Scholar], [https://github.com/zenogantner/ GitHub], [http://stackoverflow.com/users/404824/zenog StackOverflow], [http://www.kaggle.com/users/15462/zenog Kaggle], [http://www.slideshare.net/zenogantner SlideShare] | [http://www.ismll.uni-hildesheim.de/personen/gantner_en.html homepage], [https://scholar.google.com/citations?user=AhVYsaoAAAAJ Google Scholar], [https://github.com/zenogantner/ GitHub], [http://stackoverflow.com/users/404824/zenog StackOverflow], [http://www.kaggle.com/users/15462/zenog Kaggle], [http://www.slideshare.net/zenogantner SlideShare] | ||
− | |||
== TODO == | == TODO == | ||
− | * | + | * page about Fashion RecSys workshop |
− | * | + | * add link to Google tutorial |
− | * extend Person template | + | * add pages about PyTorch and TF recommendations |
− | * extend/create dataset template | + | * marker templates for sequential recommendations, embeddings, e-commerce, CTR prediction, reinforcement learning, cold-start |
+ | * extend Person template: Google Scholar, LinkedIn, SlideShare, and GitHub | ||
+ | * extend/create dataset template (link to downloads, Google scholar search, Papers with Code) | ||
+ | * event/conference template (individual events and conference series) | ||
+ | * create company template (similar to persons), with github link, tech blog, Wikipedia link, corporate page, etc. | ||
+ | * page about [[Recsperts podcast]] | ||
== Article wishlist == | == Article wishlist == | ||
* [[A/B testing]] | * [[A/B testing]] | ||
* <s>[[active learning]]</s> | * <s>[[active learning]]</s> | ||
+ | * [[approximate nearest neighbor search]] | ||
* [[attribute-aware recommendation]] | * [[attribute-aware recommendation]] | ||
* [[attribute-based recommendation]] [http://www.theatlantic.com/technology/archive/2014/01/how-netflix-reverse-engineered-hollywood/282679/] | * [[attribute-based recommendation]] [http://www.theatlantic.com/technology/archive/2014/01/how-netflix-reverse-engineered-hollywood/282679/] | ||
+ | * [[autoencoder]] | ||
+ | * [[bag-of-items]] | ||
* [[bagging]] | * [[bagging]] | ||
* [[bandit]] (-> [[multi-arm bandit]]) | * [[bandit]] (-> [[multi-arm bandit]]) | ||
Line 33: | Line 36: | ||
* [[click stream]] | * [[click stream]] | ||
* [[client-side recommendation]] (ask Chris) | * [[client-side recommendation]] (ask Chris) | ||
− | |||
* [[code recommendation]] [http://t.co/QakdUh02] | * [[code recommendation]] [http://t.co/QakdUh02] | ||
* [[CofiRank]] (ask Markus) | * [[CofiRank]] (ask Markus) | ||
Line 45: | Line 47: | ||
* [[data analytics]] | * [[data analytics]] | ||
* [[data mining]] | * [[data mining]] | ||
− | * [[decision theory]] | + | * [[decision theory]] |
+ | * [[deep learning]] | ||
* [[distance]] | * [[distance]] | ||
* [[distributed computing]] (ask Sebastian) | * [[distributed computing]] (ask Sebastian) | ||
Line 57: | Line 60: | ||
* [[FAQ for recommender system developers]] | * [[FAQ for recommender system developers]] | ||
* [[FAQ for recommender system users]] | * [[FAQ for recommender system users]] | ||
+ | * [[Fashion recommendation]], [[Fashion recommendations]] | ||
* [[Filter bubble]] (ask Alan and Neal) | * [[Filter bubble]] (ask Alan and Neal) | ||
* [[Flixster dataset]] | * [[Flixster dataset]] | ||
Line 103: | Line 107: | ||
* [[Matchbox]] [http://research.microsoft.com/en-us/um/cambridge/projects/infernet/docs/Recommender%20System.aspx] (ask Noam) | * [[Matchbox]] [http://research.microsoft.com/en-us/um/cambridge/projects/infernet/docs/Recommender%20System.aspx] (ask Noam) | ||
* <s>[[matrix factorization]]</s> | * <s>[[matrix factorization]]</s> | ||
− | * [[maximum a-priori estimation]] ([[MAP]]) | + | * [[maximum a-priori estimation]] ([[MAP]]) |
+ | * [[maximum inner product search]] | ||
* [[mean average precision]] ([[MAP]]) - link to [http://en.wikipedia.org/wiki/Information_retrieval#Mean_average_precision] | * [[mean average precision]] ([[MAP]]) - link to [http://en.wikipedia.org/wiki/Information_retrieval#Mean_average_precision] | ||
* [[mean reciprocal rank]] | * [[mean reciprocal rank]] | ||
Line 109: | Line 114: | ||
* <s>[[Million Song Dataset Challenge]]</s> (<s>ask Brian McFee</s>) | * <s>[[Million Song Dataset Challenge]]</s> (<s>ask Brian McFee</s>) | ||
* [[MinHash]] | * [[MinHash]] | ||
+ | * [[MLOps] | ||
* [[model]] | * [[model]] | ||
* [[monetization]] | * [[monetization]] | ||
Line 118: | Line 124: | ||
* <s>[[MyMedia]]</s> (thank you Alan!) | * <s>[[MyMedia]]</s> (thank you Alan!) | ||
* <s>[[NDCG]]</s> | * <s>[[NDCG]]</s> | ||
+ | * [[neural networks]] | ||
* [[news recommendation]] | * [[news recommendation]] | ||
* [[offline experiment]] | * [[offline experiment]] | ||
* [[one-class feedback]] | * [[one-class feedback]] | ||
* [[overfitting]] | * [[overfitting]] | ||
+ | * [[page composition]] | ||
* [[pairwise interaction tensor factorization]] ([[PITF]], ask Steffen) | * [[pairwise interaction tensor factorization]] ([[PITF]], ask Steffen) | ||
+ | * [[Papers with Code]] | ||
* [[parallel factor analysis]] ([[PARAFAC]]), [[canonical decomposition]] (ask Steffen) | * [[parallel factor analysis]] ([[PARAFAC]]), [[canonical decomposition]] (ask Steffen) | ||
* [[parallel matrix factorization]] | * [[parallel matrix factorization]] | ||
Line 151: | Line 160: | ||
* [[scalability]] (ask Sebastian) | * [[scalability]] (ask Sebastian) | ||
* [[semi-supervised learning]] | * [[semi-supervised learning]] | ||
+ | * [[sequential recommendation]] | ||
* [[serendipity]] (ask Alan, ask Ben) | * [[serendipity]] (ask Alan, ask Ben) | ||
+ | * [[session-based recommendation]] | ||
* [[similarity]] | * [[similarity]] | ||
* [[SmartTypes]] [http://www.smarttypes.org/blog/graphlab_datasets] | * [[SmartTypes]] [http://www.smarttypes.org/blog/graphlab_datasets] | ||
Line 163: | Line 174: | ||
* [[TaFeng]] | * [[TaFeng]] | ||
* <s>[[tag]]</s> (thanks Alan) | * <s>[[tag]]</s> (thanks Alan) | ||
− | |||
* [[Tanimoto coefficient]] --> [[Jaccard index]] | * [[Tanimoto coefficient]] --> [[Jaccard index]] | ||
* [[Tapestry]] | * [[Tapestry]] | ||
Line 173: | Line 183: | ||
* [[Tucker decomposition]] (ask Steffen) | * [[Tucker decomposition]] (ask Steffen) | ||
* [[TV program recommendation]] (ask Chris) | * [[TV program recommendation]] (ask Chris) | ||
− | * [[UMAP | + | * [[UMAP]] |
* [[user]] | * [[user]] | ||
* [[user-item matrix]] | * [[user-item matrix]] | ||
Line 181: | Line 191: | ||
* [[user satisfaction]] | * [[user satisfaction]] | ||
* [[video recommendation]] | * [[video recommendation]] | ||
− | * | + | * [[WSDM]] |
− | + | * [[Yahoo Movie Dataset]] | |
− | |||
− | |||
− | |||
− | |||
− | * [[Yahoo Movie Dataset]] | ||
=== RecSys people === | === RecSys people === | ||
+ | |||
* [[Joseph Konstan]] | * [[Joseph Konstan]] | ||
* [[John Riedl]] | * [[John Riedl]] | ||
Line 196: | Line 202: | ||
* [[Greg Linden]] | * [[Greg Linden]] | ||
* [[Paul Lamere]] | * [[Paul Lamere]] | ||
− | * | + | * [[Ted Dunning]] |
+ | * [[Sebastian Schelter]] -- https://scholar.google.de/citations?user=zCpQUukAAAAJ&hl=en -- https://github.com/sscdotopen -- https://github.com/schelterlabs | ||
+ | * [[Ralf Herbrich]] | ||
=== Companies === | === Companies === | ||
− | |||
* [[Alleyoop]] -- [http://gigaom.com/2012/06/06/with-new-recommendation-engine-alleyoop-wants-to-be-a-tutor-for-teens] | * [[Alleyoop]] -- [http://gigaom.com/2012/06/06/with-new-recommendation-engine-alleyoop-wants-to-be-a-tutor-for-teens] | ||
− | * [[ | + | * [[Baidu]] -- [http://www.yichang-cs.com/jlu/WSDM21_unbiased.pdf] |
− | |||
* [[BBC]] -- [http://www.bbc.co.uk/blogs/researchanddevelopment/2012/05/client-side-recommendations.shtml] | * [[BBC]] -- [http://www.bbc.co.uk/blogs/researchanddevelopment/2012/05/client-side-recommendations.shtml] | ||
− | * [[BMAT]] ( | + | * [[BMAT]] |
− | * [[Commendo]] | + | * [[Bol.com]] |
+ | * [[Booking.com]] [https://dl.acm.org/doi/pdf/10.1145/3383313.3412215] [https://www.researchgate.net/profile/Dmitri-Goldenberg/publication/349762238_Personalization_in_Practice_Methods_and_Applications/links/60409bf04585154e8c75323d/Personalization-in-Practice-Methods-and-Applications.pdf] [https://dl.acm.org/doi/pdf/10.1145/3383313.3412215] [https://dl.acm.org/doi/abs/10.1145/3460231.3474611] [https://dl.acm.org/doi/abs/10.1145/3511808.3557100] [https://dl.acm.org/doi/abs/10.1145/3292500.3330744] [http://www.toinebogers.com/workshops/complexrec2020/Mavridis.pdf] [https://arxiv.org/abs/2109.06723] [https://dl.acm.org/doi/abs/10.1145/3460231.3474611] | ||
+ | * '''[[ByteDance]]''' and [[TikTok]]/[[Douyin]] (redirect), [https://newsroom.tiktok.com/en-us/how-tiktok-recommends-videos-for-you], [https://github.com/bytedance/Hammer], [https://github.com/bytedance/byteps], [https://github.com/bytedance/LargeBatchCTR], [https://dl.acm.org/doi/pdf/10.1145/3308558.3313447] [https://arxiv.org/pdf/2204.06240.pdf] | ||
+ | * [[Commendo]] | ||
* [[Directed Edge]] -- http://www.directededge.com | * [[Directed Edge]] -- http://www.directededge.com | ||
* [[EBay]] | * [[EBay]] | ||
− | * [[The Echo Nest]] [http://blog.echonest.com/post/33229165293/taste-profiles-go-public] [http://notes.variogr.am/post/37675885491/how-music-recommendation-works-and-doesnt-work] | + | * [[The Echo Nest]] [http://blog.echonest.com/post/33229165293/taste-profiles-go-public] [http://notes.variogr.am/post/37675885491/how-music-recommendation-works-and-doesnt-work] => [[Spotify]] |
− | + | * [[Etsy]] | |
− | + | * [[foursquare]] -- [http://engineering.foursquare.com/2011/03/22/building-a-recommendation-engine-foursquare-style/] [http://engineering.foursquare.com/2012/03/23/machine-learning-with-large-networks-of-people-and-places/] | |
− | + | * [[Froomle]] | |
− | * | ||
− | * [[foursquare]] -- [http://engineering.foursquare.com/2011/03/22/building-a-recommendation-engine-foursquare-style/] [http://engineering.foursquare.com/2012/03/23/machine-learning-with-large-networks-of-people-and-places/] | ||
− | * [[ | ||
− | |||
− | |||
− | |||
− | |||
* [[Hunch]] | * [[Hunch]] | ||
+ | * [[Ikea]] | ||
* [[Kaggle]] | * [[Kaggle]] | ||
− | * | + | * [[Kuaishou]] [https://en.wikipedia.org/wiki/Kuaishou], [https://arxiv.org/abs/2302.01724], [https://arxiv.org/pdf/2212.02779.pdf] |
* [[last.fm]] -- [http://www.slideshare.net/MarkLevy/algorithms-on-hadoop-at-lastfm] [http://www.quora.com/How-does-Last-fm-compute-lists-of-similar-artists] | * [[last.fm]] -- [http://www.slideshare.net/MarkLevy/algorithms-on-hadoop-at-lastfm] [http://www.quora.com/How-does-Last-fm-compute-lists-of-similar-artists] | ||
− | |||
* [[Lumi]] | * [[Lumi]] | ||
− | * [[Microsoft]] | + | * [[Microsoft]] |
− | + | * [[Myrrix]] | |
− | * [[Myrrix]] | + | * [[Nokia]] -- add 2011 Buzzwords presentation |
− | * [[ | + | * [[Otto]] |
− | * [[ | ||
* [[outbrain]] -- [http://www.conversationagent.com/2012/05/content-recommendation-engine-outbrain.html] | * [[outbrain]] -- [http://www.conversationagent.com/2012/05/content-recommendation-engine-outbrain.html] | ||
* [[Pandora]] [http://www.fastcompany.com/1808123/tom-conrad-pandora-music-genome-project] [http://blog.pandora.com/pandora/archives/2012/10/pandora-and-art.html] (ask Tao) | * [[Pandora]] [http://www.fastcompany.com/1808123/tom-conrad-pandora-music-genome-project] [http://blog.pandora.com/pandora/archives/2012/10/pandora-and-art.html] (ask Tao) | ||
− | * [[ | + | * [[Pinterest]] -- [https://arxiv.org/pdf/1702.07969.pdf] |
* [[Prudsys]] | * [[Prudsys]] | ||
* [[Recommind]] [http://www.recommind.com/] | * [[Recommind]] [http://www.recommind.com/] | ||
* [[RichRelevance]] (ask Darren) | * [[RichRelevance]] (ask Darren) | ||
* [[Samsung]] | * [[Samsung]] | ||
− | |||
* [[sematext]] | * [[sematext]] | ||
+ | * [[ShareChat]] -- [https://dl.acm.org/doi/abs/10.1145/3543873.3587679] | ||
+ | * [[Shopify]] -- ACM RecSys | ||
* [[Sidebar]] | * [[Sidebar]] | ||
− | * [[SoundCloud]] | + | * [[SoundCloud]] |
− | * [[Spotify]] -- [http://paidcontent.org/2012/12/06/spotify-solves-discovery-by-discovering-music-aint-so-social-after-all/] [http://vimeo.com/57900625] | + | * [[Spotify]] -- [http://paidcontent.org/2012/12/06/spotify-solves-discovery-by-discovering-music-aint-so-social-after-all/] [http://vimeo.com/57900625] [https://urldefense.com/v3/__https://www.wsj.com/video/series/wsj-explains/how-spotify-knows-what-you-want-to-hear-next/E91EB935-C3EE-42FF-B41A-246614F8F1A1?mod=tech_lead_pos5__;!!Bt8RZUm9aw!9jWj6_eMhtjp9bJ_vhQae7R843i0wMfn2_BVASWQvcyzT13IedsQDJaJiyNBokkz4L2UKM0H4C8I8s4T$] |
* [[Strands]] | * [[Strands]] | ||
* [[TiVo]] | * [[TiVo]] | ||
− | * [[Twitter]] [http://engineering.twitter.com/2012/03/generating-recommendations-with.html] | + | * [[Twitter]] [http://engineering.twitter.com/2012/03/generating-recommendations-with.html] [https://blog.twitter.com/engineering/en_us/topics/open-source/2023/twitter-recommendation-algorithm] [https://github.com/twitter/the-algorithm] [https://github.com/twitter/sbf] |
− | * [[Yahoo]] | + | * '''[[Yahoo]]''' [https://people.csail.mit.edu/romer/papers/TISTRespPredAds.pdf] |
− | * [[ | + | * [[Yandex]] [https://openreview.net/pdf?id=MXfTQp8bZF] |
− | * [[ | + | * [[YooChoose]] |
* [[Zite]] | * [[Zite]] | ||
Latest revision as of 06:39, 28 September 2023
Zeno Gantner, formerly at University of Hildesheim, Germany. Now working at Zalando in Berlin. Primary developer of the MyMediaLite recommender system library. Co-organizer of the Recommender Stammtisch in Berlin.
homepage, Google Scholar, GitHub, StackOverflow, Kaggle, SlideShare
TODO
- page about Fashion RecSys workshop
- add link to Google tutorial
- add pages about PyTorch and TF recommendations
- marker templates for sequential recommendations, embeddings, e-commerce, CTR prediction, reinforcement learning, cold-start
- extend Person template: Google Scholar, LinkedIn, SlideShare, and GitHub
- extend/create dataset template (link to downloads, Google scholar search, Papers with Code)
- event/conference template (individual events and conference series)
- create company template (similar to persons), with github link, tech blog, Wikipedia link, corporate page, etc.
- page about Recsperts podcast
Article wishlist
- A/B testing
active learning- approximate nearest neighbor search
- attribute-aware recommendation
- attribute-based recommendation [1]
- autoencoder
- bag-of-items
- bagging
- bandit (-> multi-arm bandit)
- beer recommendation -- very important task ... (ask Ben)
blogs- BookCrossing (ask Cai-Nicolas)
- capped binomial deviation (CBD)
- Category:File format
- CHI (ask Alan)
- choice overload (ask Bart, Martijn, Dirk)
- click stream
- client-side recommendation (ask Chris)
- code recommendation [2]
- CofiRank (ask Markus)
cold-start problem- computational advertising
content-based filteringcontextcontext-aware recommendation- contextual bandit
- cross-validation [3]
- data analytics
- data mining
- decision theory
- deep learning
- distance
- distributed computing (ask Sebastian)
- distributed matrix factorization (ask Rainer)
- Eigentaste
- Epinions dataset
- Explanations (ask Nava)
- exploration vs. exploitation
- evaluation
- factorization model, factorization models
- FAQ for recommender system developers
- FAQ for recommender system users
- Fashion recommendation, Fashion recommendations
- Filter bubble (ask Alan and Neal)
- Flixster dataset
- F measure, F1 measure
- fold-in [4]
- GraphChi (ask Danny)
- GraphLab (ask Danny)
- Greg Linden
- grid search [5]
group recommendationHarry Potter effect- HCI
- higher-order SVD (ask Steffen)
hybrid recommendation- hyperparameter
- incentive
- Infer.NET [6]
- information retrieval
- Introduction to recommender systems
- Introduction to recommender system algorithms
- IPTV (ask Chris)
- item
- IUI: IUI 2010, IUI 2011, IUI 2012, IUI 2013
- Jaccard index
- Jester
- job recommendation
- Joke recommendation
- KDD Cup: KDD Cup 2010 KDD Cup 2011 KDD Cup 2012
- KDD: KDD 2007, KDD 2008, KDD 2009, KDD 2010
- keyword-based recommendation
kNN- lab testing
- latency (ask Sebastian)
- latent factor model
- learning
- learning to rank
- List of acronyms -- cmp. http://aclweb.org/aclwiki/index.php?title=Acronyms
- List of recommender system meetings
- live evaluation (ask Andreas H./Alan)
- location-aware recommendation
- London RecSys Meetup (ask Neal)
- long tail (ask Oscar)
- machine learning
- Markov chain (ask Christoph)
- Markov decision process, MDP
- Matchbox [7] (ask Noam)
matrix factorization- maximum a-priori estimation (MAP)
- maximum inner product search
- mean average precision (MAP) - link to [8]
- mean reciprocal rank
Million Song DatasetMillion Song Dataset Challenge(ask Brian McFee)- MinHash
- [[MLOps]
- model
- monetization
- Movie Hack Day (ask Jannis and Alan)
- multi-arm bandit (ask Matt)
- Music Hack Day (ask Amelie)
- music information retrieval (ask Oscar, Ben, Amelie, Markus)
music recommendationMyMedia(thank you Alan!)NDCG- neural networks
- news recommendation
- offline experiment
- one-class feedback
- overfitting
- page composition
- pairwise interaction tensor factorization (PITF, ask Steffen)
- Papers with Code
- parallel factor analysis (PARAFAC), canonical decomposition (ask Steffen)
- parallel matrix factorization
- parameter
Pearson correlation- personalization
- personalized advertising
- personalized prices [9]
- personalized search
- positive-only feedback
- preference elicitation (ask Martijn and Bart)
- product recommendation
- public transport (ask Neal)
- R
- ranking
- RecDB (ask, http://www-users.cs.umn.edu/~sarwat/RecDB/)
- recipe recommendation
- recommendation of financial products
- recommender lab (ask Michael H.)
recommender system- RecSys meetups (do it yourself)
- reinforcement learning (ask Tobias)
regularization- reputation
- restricted Boltzmann machine (ask Andriy)
- review
- Ringo
- scalability (ask Sebastian)
- semi-supervised learning
- sequential recommendation
- serendipity (ask Alan, ask Ben)
- session-based recommendation
- similarity
- SmartTypes [10]
- software as a service (ask Manuel B.)
- software recommendation
- standard benchmarks TODO
- state of the art cmp. http://aclweb.org/aclwiki/index.php?title=State_of_the_art
- stream processing
SVDSVD++, SVDPlusPlus- TaFeng
tag(thanks Alan)- Tanimoto coefficient --> Jaccard index
- Tapestry
- tensor factorization (ask Steffen)
- text-based recommendation
- text mining
- time-aware recommendation
- transductive learning
- Tucker decomposition (ask Steffen)
- TV program recommendation (ask Chris)
- UMAP
- user
- user-item matrix
- user model
- user preferences
- user recommendation
- user satisfaction
- video recommendation
- WSDM
- Yahoo Movie Dataset
RecSys people
- Joseph Konstan
- John Riedl
- Yehuda Koren
- Pearl Pu
- Greg Linden
- Paul Lamere
- Ted Dunning
- Sebastian Schelter -- https://scholar.google.de/citations?user=zCpQUukAAAAJ&hl=en -- https://github.com/sscdotopen -- https://github.com/schelterlabs
- Ralf Herbrich
Companies
- Alleyoop -- [11]
- Baidu -- [12]
- BBC -- [13]
- BMAT
- Bol.com
- Booking.com [14] [15] [16] [17] [18] [19] [20] [21] [22]
- ByteDance and TikTok/Douyin (redirect), [23], [24], [25], [26], [27] [28]
- Commendo
- Directed Edge -- http://www.directededge.com
- EBay
- The Echo Nest [29] [30] => Spotify
- Etsy
- foursquare -- [31] [32]
- Froomle
- Hunch
- Ikea
- Kaggle
- Kuaishou [33], [34], [35]
- last.fm -- [36] [37]
- Lumi
- Microsoft
- Myrrix
- Nokia -- add 2011 Buzzwords presentation
- Otto
- outbrain -- [38]
- Pandora [39] [40] (ask Tao)
- Pinterest -- [41]
- Prudsys
- Recommind [42]
- RichRelevance (ask Darren)
- Samsung
- sematext
- ShareChat -- [43]
- Shopify -- ACM RecSys
- Sidebar
- SoundCloud
- Spotify -- [44] [45] [46]
- Strands
- TiVo
- Twitter [47] [48] [49] [50]
- Yahoo [51]
- Yandex [52]
- YooChoose
- Zite
RecSys slides, classes, etc.
- http://www.lsi.dsc.ufcg.edu.br/lib/exe/fetch.php?id=bd_lanche&cache=cache&media=fatoracao_matrizes.pdf
- Berkeley: Practical Machine Learning: collaborative filtering (only rating prediction)
- http://alex.smola.org/teaching/berkeley2012/recommender.html
- http://cms.uni-konstanz.de/informatik/rendle/teaching/ss2012/fm0/