Abstract:Factorization Machines (FMs) are effective in incorporating side information to overcome the cold-start and data sparsity problems in recommender systems. Traditional FMs adopt the inner product to model the second-order interactions between different attributes, which are represented via feature vectors. The problem is that the inner product violates the triangle inequality property of feature vectors. As a result, it cannot well capture fine-grained attribute interactions, resulting in sub-optimal performance. Recently, the euclidean distance is exploited in FMs to replace the inner product and has delivered better performance. However, previous FM methods including the ones equipped with the euclidean distance all focus on the attribute-level interaction modeling, ignoring the critical intrinsic feature correlations inside attributes. Thereby, they fail to model the complex and rich interactions exhibited in the real-world data. To tackle this problem, in this paper, we propose a FM framework equipped with generalized metric learning techniques to better capture these feature correlations. In particular, based on this framework, we present a Mahalanobis distance and a deep neural network (DNN) methods, which can effectively model the linear and non-linear correlations between features, respectively. Besides, we design an efficient approach for simplifying the model functions. Experiments on several benchmark datasets demonstrate that our proposed framework outperforms several state-of-the-art baselines by a large margin. Moreover, we collect a new large-scale dataset on second-hand trading to justify the effectiveness of our method over cold-start and data sparsity problems in recommender systems.

Locally Linear Factorization Machines.

L G ] 1 7 A pr 2 01 8 A Boosting Framework of Factorization Machine

A Boosting Framework of Factorization Machine.

Robust Factorization Machine - A Doubly Capped Norms Minimization.

Convex Factorization Machine for Regression

Sketched Follow-The-Regularized-Leader for Online Factorization Machine.

Unified Locally Linear Classifiers with Diversity-Promoting Anchor Points

Attention-over-Attention Field-Aware Factorization Machine.

Online Compact Convexified Factorization Machine.

Feature Fusion Using Locally Linear Embedding for Classification

Local Feature Discriminant Projection

Enhancing Factorization Machines with Generalized Metric Learning

Local Fisher Discriminant Analysis with Locally Linear Embedding Affinity Matrix

Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks

FSLLE: A Fast K Selection Algorithm for Locally Linear Embedding.

Neural Factorization Machines for Sparse Predictive Analytics

Locally Linear Representation Fisher Criterion

Local Learning Regularized Nonnegative Matrix Factorization.

Fisher'S Linear Discriminant (fld) and Support Vector Machine (svm) in Non-Negative Matrix Factorization (nmf) Residual Space for Face Recognition

Enhancing Factorization Machines with Generalized Metric Learning (extended Abstract).