Jarana Manotumruksa

Matrix Factorisation with Word Embeddings for Rating Prediction on Location-Based Social Networks

With vast amounts of data being created on location-based social networks (LBSNs) such as Yelp and Foursquare, making effective personalised suggestions to users is an essential functionality. Matrix Factorisation (MF) is a collaborative filtering-based approach that is widely used to generate suggestions relevant to user’s preferences. In this paper, we address the problem of predicting the rating that users give to venues they visit. Previous works have proposed MF-based approaches that consider auxiliary information (e.g. social information and users’ comments on venues) to improve the accuracy of rating predictions. Such approaches leverage the users’ friends’ preferences, extracted from either ratings or comments, to regularise the complexity of MF-based models and to avoid over-fitting. However, social information may not be available, e.g. due to privacy concerns. To overcome this limitation, in this paper, we propose a novel MF-based approach that exploits word embeddings to effectively model users’ preferences and the characteristics of venues from the textual content of comments left by users, regardless of their relationship. Experiments conducted on a large dataset of LBSN ratings demonstrate the effectiveness of our proposed approach compared to various state-of-the-art rating prediction approaches.

Predicting Contextually Appropriate Venues in Location-Based Social Networks

The effective suggestion of venues that are appropriate for a user to visit is a challenging problem, as the appropriateness of a venue can depend on particular contextual aspects, such as the duration of the user’s visit, or the composition of the user’s travelling group (e.g. alone, with friends, or with family). This paper proposes a supervised approach that predicts appropriateness of venues to particular contextual aspects, by leveraging user-generated data in Location-Based Social Networks (LBSNs) such as Foursquare. Our approach learns a binary classifier for each dimension of three considered contextual aspects. A set of discriminative features are extracted from the comments, photos and website of venues. Using a dataset from the TREC 2015 Contextual Suggestion track, supplemented by venue annotations generated by crowdsourcing, we conduct a comprehensive experimental study to identify the set of features appropriate for our problem and to evaluate the effectiveness of our proposed approach. Our results demonstrate both the accuracy of our classification approach in predicting suitable contextual aspects for a venue, and its effectiveness at making better venue recommendations than the best performing system in TREC 2015.

A Deep Recurrent Collaborative Filtering Framework for Venue Recommendation

Venue recommendation is an important application for Location-Based Social Networks (LBSNs), such as Yelp, and has been extensively studied in recent years. Matrix Factorisation (MF) is a popular Collaborative Filtering (CF) technique that can suggest relevant venues to users based on an assumption that similar users are likely to visit similar venues. In recent years, deep neural networks have been successfully applied to tasks such as speech recognition, computer vision and natural language processing. Building upon this momentum, various approaches for recommendation have been proposed in the literature to enhance the effectiveness of MF-based approaches by exploiting neural network models such as: word embeddings to incorporate auxiliary information (e.g. textual content of comments); and Recurrent Neural Networks (RNN) to capture sequential properties of observed user-venue interactions. However, such approaches rely on the traditional inner product of the latent factors of users and venues to capture the concept of collaborative filtering, which may not be sufficient to capture the complex structure of user-venue interactions. In this paper, we propose a Deep Recurrent Collaborative Filtering framework (DRCF) with a pairwise ranking function that aims to capture user-venue interactions in a CF manner from sequences of observed feedback by leveraging Multi-Layer Perception and Recurrent Neural Network architectures. Our proposed framework consists of two components: namely Generalised Recurrent Matrix Factorisation (GRMF) and Multi-Level Recurrent Perceptron (MLRP) models. In particular, GRMF and MLRP learn to model complex structures of user-venue interactions using element-wise and dot products as well as the concatenation of latent factors. In addition, we propose a novel sequence-based negative sampling approach that accounts for the sequential properties of observed feedback and geographical location of venues to enhance the quality of venue suggestions, as well as alleviate the cold-start users problem. Experiments on three large checkin and rating datasets show the effectiveness of our proposed framework by outperforming various state-of-the-art approaches.

A Personalised Ranking Framework with Multiple Sampling Criteria for Venue Recommendation

Recommending a ranked list of interesting venues to users based on their preferences has become a key functionality in Location-Based Social Networks (LBSNs) such as Yelp and Gowalla. Bayesian Personalised Ranking (BPR) is a popular pairwise recommendation technique that is used to generate the ranked list of venues of interest to a user, by leveraging the users implicit feedback such as their check-ins as instances of positive feedback, while randomly sampling other venues as negative instances. To alleviate the sparsity that affects the usefulness of recommendations by BPR for users with few check-ins, various approaches have been proposed in the literature to incorporate additional sources of information such as the social links between users, the textual content of comments, as well as the geographical location of the venues. However, such approaches can only readily leverage one source of additional information for negative sampling. Instead, we propose a novel Personalised Ranking Framework with Multiple sampling Criteria (PRFMC) that leverages both geographical influence and social correlation to enhance the effectiveness of BPR. In particular, we apply a multi-centre Gaussian model and a power-law distribution method, to capture geographical influence and social correlation when sampling negative venues, respectively. Finally, we conduct comprehensive experiments using three large-scale datasets from the Yelp, Gowalla and Brightkite LBSNs. The experimental results demonstrate the effectiveness of fusing both geographical influence and social correlation in our proposed PRFMC framework and its superiority in comparison to BPR-based and other similar ranking approaches. Indeed, our PRFMC approach attains a 37% improvement in MRR over a recently proposed approach that identifies negative venues only from social links.

Searching the Internet of Things

Description: Internet-of-Things (IoT) Analytics are an integral element of most IoT applications, as it provides the means to extract knowledge, drive actuation services and optimize decision making. IoT analytics will be a major contributor to IoT business value in the coming years, as it will enable organizations to process and fully leverage large amounts of IoT data, which are nowadays largely underutilized.