In recent years, deep neural networks have yielded immense success on speech recognition, computer vision and natural language processing. However, the exploration of deep neural networks on recommender systems has received relatively less scrutiny. In this work, we strive to develop techniques based on neural networks to tackle the key problem in recommendation --- collaborative filtering --- on the basis of implicit feedback. Although some recent work has employed deep learning for recommendation, they primarily used it to model auxiliary information, such as textual descriptions of items and acoustic features of musics. When it comes to model the key factor in collaborative filtering --- the interaction between user and item features, they still resorted to matrix factorization and applied an inner product on the latent features of users and items. By replacing the inner product with a neural architecture that can learn an arbitrary function from data, we present a general framework named NCF, short for Neural network-based Collaborative Filtering. NCF is generic and can express and generalize matrix factorization under its framework. To supercharge NCF modelling with non-linearities, we propose to leverage a multi-layer perceptron to learn the user-item interaction function. Extensive experiments on two real-world datasets show significant improvements of our proposed NCF framework over the state-of-the-art methods. Empirical evidence shows that using deeper layers of neural networks offers better recommendation performance.

Neural Collaborative Filtering

In the last years, wireless sensor networks (WSNs) have gained increasing attention from both the research community and actual users. As sensor nodes are generally battery-powered devices, the critical aspects to face concern how to reduce the energy consumption of nodes, so that the network lifetime can be extended to reasonable times. In this paper we first break down the energy consumption for the components of a typical sensor node, and discuss the main directions to energy conservation in WSNs. Then, we present a systematic and comprehensive taxonomy of the energy conservation schemes, which are subsequently discussed in depth. Special attention has been devoted to promising solutions which have not yet obtained a wide attention in the literature, such as techniques for energy efficient data acquisition. Finally we conclude the paper with insights for research directions about energy conservation in WSNs.

/pdf/energy-conservation-in-wireless-sensor-networks-a-survey-1bl7d2obnm.pdf

Energy conservation in wireless sensor networks: A survey

Matrix Factorization Techniques for Recommender Systems

When building a unified vision system or gradually adding new capabilities to a system, the usual assumption is that training data for all tasks is always available. However, as the number of tasks grows, storing and retraining on such data becomes infeasible. A new problem arises where we add new capabilities to a Convolutional Neural Network (CNN), but the training data for its existing capabilities are unavailable. We propose our Learning without Forgetting method, which uses only new task data to train the network while preserving the original capabilities. Our method performs favorably compared to commonly used feature extraction and fine-tuning adaption techniques and performs similarly to multitask learning that uses original task data we assume unavailable. A more surprising observation is that Learning without Forgetting may be able to replace fine-tuning with similar old and new task datasets for improved new task performance.

Learning without Forgetting

Many predictive tasks of web applications need to model categorical variables, such as user IDs and demographics like genders and occupations. To apply standard machine learning techniques, these categorical predictors are always converted to a set of binary features via one-hot encoding, making the resultant feature vector highly sparse. To learn from such sparse data effectively, it is crucial to account for the interactions between features. Factorization Machines (FMs) are a popular solution for efficiently using the second-order feature interactions. However, FM models feature interactions in a linear way, which can be insufficient for capturing the non-linear and complex inherent structure of real-world data. While deep neural networks have recently been applied to learn non-linear feature interactions in industry, such as the Wide&Deep by Google and DeepCross by Microsoft, the deep structure meanwhile makes them difficult to train. In this paper, we propose a novel model Neural Factorization Machine (NFM) for prediction under sparse settings. NFM seamlessly combines the linearity of FM in modelling second-order feature interactions and the non-linearity of neural network in modelling higher-order feature interactions. Conceptually, NFM is more expressive than FM since FM can be seen as a special case of NFM without hidden layers. Empirical results on two regression tasks show that with one hidden layer only, NFM significantly outperforms FM with a 7.3% relative improvement. Compared to the recent deep learning methods Wide&Deep and DeepCross, our NFM uses a shallower structure but offers better performance, being much easier to train and tune in practice.

Neural Factorization Machines for Sparse Predictive Analytics

In recent years, interest in recommender research has shifted from explicit feedback towards implicit feedback data. A diversity of complex models has been proposed for a wide variety of applications. Despite this, learning from implicit feedback is still computationally challenging. So far, most work relies on stochastic gradient descent (SGD) solvers which are easy to derive, but in practice challenging to apply, especially for tasks with many items. For the simple matrix factorization model, an efficient coordinate descent (CD) solver has been previously proposed. However, efficient CD approaches have not been derived for more complex models. In this paper, we provide a new framework for deriving efficient CD algorithms for complex recommender models. We identify and introduce the property of k-separable models. We show that k-separability is a sufficient property to allow efficient optimization of implicit recommender problems with CD. We illustrate this framework on a variety of state-of-the-art models including factorization machines and Tucker decomposition. To summarize, our work provides the theory and building blocks to derive efficient implicit CD algorithms for complex recommender models.

A Generic Coordinate Descent Framework for Learning from Implicit Feedback

In this paper, we address the problem of extending a relational database system to facilitate efficient real-time application of dynamic probabilistic models to streaming data. We use the recently proposed abstraction of model-based views for this purpose, by allowing users to declaratively specify the model to be applied, and by presenting the output of the models to the user as a probabilistic database view. We support declarative querying over such views using an extended version of SQL that allows for querying probabilistic data. Underneath we use particle filters, a class of sequential Monte Carlo algorithms, to represent the present and historical states of the model as sets of weighted samples (particles) that are kept up-to-date as new data arrives. We develop novel techniques to convert the queries on the model-based view directly into queries over particle tables, enabling highly efficient query processing. Finally, we present experimental evaluation of our prototype implementation over several synthetic and real datasets, that demonstrates the feasibility of online modeling of streaming data using our system and establishes the advantages of tight integration between dynamic probabilistic models and databases.

/pdf/online-filtering-smoothing-and-probabilistic-modeling-of-25e57in22b.pdf

Online Filtering, Smoothing and Probabilistic Modeling of Streaming data

Recommender systems based on latent factor models have been effectively used for understanding user interests and predicting future actions. Such models work by projecting the users and items into a smaller dimensional space, thereby clustering similar users and items together and subsequently compute similarity between unknown user-item pairs. When user-item interactions are sparse (sparsity problem) or when new items continuously appear (cold start problem), these models perform poorly. In this paper, we exploit the combination of taxonomies and latent factor models to mitigate these issues and improve recommendation accuracy. We observe that taxonomies provide structure similar to that of a latent factor model: namely, it imposes human-labeled categories (clusters) over items. This leads to our proposed taxonomy-aware latent factor model (TF) which combines taxonomies and latent factors using additive models. We develop efficient algorithms to train the TF models, which scales to large number of users/items and develop scalable inference/recommendation algorithms by exploiting the structure of the taxonomy. In addition, we extend the TF model to account for the temporal dynamics of user interests using high-order Markov chains. To deal with large-scale data, we develop a parallel multi-core implementation of our TF model. We empirically evaluate the TF model for the task of predicting user purchases using a real-world shopping dataset spanning more than a million users and products. Our experiments demonstrate the benefits of using our TF models over existing approaches, in terms of both prediction accuracy and running time.

/pdf/supercharging-recommender-systems-using-taxonomies-for-163ss8jnsp.pdf

Supercharging recommender systems using taxonomies for learning user purchase behavior

Probabilistic database systems have successfully established themselves as a tool for managing uncertain data. However, much of the research in this area has focused on efficient query evaluation and has largely ignored two key issues that commonly arise in uncertain data management: First, how to provide explanations for query results, e.g., Why is this tuple in my result? or Why does this output tuple have such high probability?. Second, the problem of determining the sensitive input tuples for the given query, e.g., users are interested to know the input tuples that can substantially alter the output, when their probabilities are modified (since they may be unsure about the input probability values). Existing systems provide the lineage/provenance of each of the output tuples in addition to the output probabilities, which is a boolean formula indicating the dependence of the output tuple on the input tuples. However, lineage does not immediately provide a quantitative relationship and it is not informative when we have multiple output tuples. In this paper, we propose a unified framework that can handle both the issues mentioned above to facilitate robust query processing. We formally define the notions of influence and explanations and provide algorithms to determine the top-l influential set of variables and the top-l set of explanations for a variety of queries, including conjunctive queries, probabilistic threshold queries, top-k queries and aggregation queries. Further, our framework naturally enables highly efficient incremental evaluation when input probabilities are modified (e.g., if uncertainty is resolved). Our preliminary experimental results demonstrate the benefits of our framework for performing robust query processing over probabilistic databases.

/pdf/sensitivity-analysis-and-explanations-for-robust-query-2o3hz7vhr3.pdf

Sensitivity analysis and explanations for robust query evaluation in probabilistic databases

With large amounts of correlated probabilistic data being generated in a wide range of application domains including sensor networks, information extraction, event detection etc., effectively managing and querying them has become an important research direction. While there is an exhaustive body of literature on querying independent probabilistic data, supporting efficient queries over large-scale, correlated databases remains a challenge. In this paper, we develop efficient data structures and indexes for supporting inference and decision support queries over such databases. Our proposed hierarchical data structure is suitable both for in-memory and disk-resident databases. We represent the correlations in the probabilistic database using a junction tree over the tuple-existence or attribute-value random variables, and use tree partitioning techniques to build an index structure over it. We show how to efficiently answer inference and aggregation queries using such an index, resulting in orders of magnitude performance benefits in most cases. In addition, we develop novel algorithms for efficiently keeping the index structure up-to-date as changes (inserts, updates) are made to the probabilistic database. We present a comprehensive experimental study illustrating the benefits of our approach to query processing in probabilistic databases.

/pdf/indexing-correlated-probabilistic-databases-2b2yo0v272.pdf

Bhargav Kanagal

Papers

A Generic Coordinate Descent Framework for Learning from Implicit Feedback

Online Filtering, Smoothing and Probabilistic Modeling of Streaming data

Supercharging recommender systems using taxonomies for learning user purchase behavior

Sensitivity analysis and explanations for robust query evaluation in probabilistic databases

Indexing correlated probabilistic databases