Ordinal Boltzmann Machines for Collaborative Filtering

By Tran The Truyen, Dinh Q. Phung, Svetha Venkatesh. 2009

72 citations recorded by the source corpus.

Collaborative filtering is an effective recommendation technique wherein the preference of an individual can potentially be predicted based on preferences of other members. Early algorithms often relied on the strong locality in the preference data, that is, it is enough to predict preference of a user on a particular item based on a small subset of other users with similar tastes or of other items with similar properties. More recently, dimensionality reduction techniques have proved to be equally competitive, and these are based on the co-occurrence patterns rather than locality. This paper explores and extends a probabilistic model known as Boltzmann Machine for collaborative filtering tasks. It seamlessly integrates both the similarity and co-occurrence in a principled manner. In particular, we study parameterisation options to deal with the ordinal nature of the preferences, and propose a joint modelling of both the user-based and item-based processes. Experiments on moderate and large-scale movie recommendation show that our framework rivals existing well-known methods.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

Neural Collaborative Filtering (2017, background): “The work was been later extended to model the ordinal nature of ratings [36].”
A Neural Autoregressive Approach to Collaborative Filtering (2016, background): “To go one step further, following Truyen et al. (2009), we take the ordinal nature of a user’s preference into consideration.”
Artificial intelligence in recommender systems (2020, methodology): “This method attracted great attention in the 2009 Netﬂix Prize competition [81], but there has been little follow-up work apart from research by Truyen et al., who extended this work by studying the parameterization options of RBM in recommendation [82].”
A review on deep learning for recommender systems: challenges and remedies (2018, methodology): “…regarding data modeling User-item preference matrix Content/side information Both Salakhutdinov et al. (2007) Oord et al. (2013) Wang et al. (2015b) Truyen et al. (2009) Wang and Wang (2014) Li et al. (2015) Georgiev and Nakov (2013) Elkahky et al. (2015) Jia et al. (2015) Sedhain et al. (2015)…”
Multi-Domain Collaborative Filtering (2010, methodology): “Many learning models have been used for CF, such as Bayesian belief networks [4], graphical models [11, 28], and dependency networks [9].”
Deep Collaborative Filtering via Marginalized Denoising Auto-encoder (2015, background): “Categories and Subject Descriptors H.2.8 [Database Management]: Data Mining; H.3.3 [Information Search Retrieval]: Information Filtering Keywords Collaborative filtering; matrix factorization; deep learning; denoising auto-encoder”
A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines (2013, methodology): “Our real-valued hybrid RBM model further outperforms the ordinal joint user-item BM-based model (ORD-UI-BM) proposed by Truyen et al. (2009). Finally, we can see in the last row of Table 2 that the additional preprocessing, which includes useruser and item-item correlation computation followed by ne”
Multitask Matrix Completion for Learning Protein Interactions Across Diseases (2016, background): “A majority of the prior work in the relevant areas of collaborative filtering and link prediction includes single relation models that use neighbourhood based prediction [21], matrix factorization based approaches [11, 14] and bayesian approaches using graphical models [10, 17].”