Collaborative Reflection-Augmented Autoencoder Network for Recommender Systems

As the deep learning techniques have expanded to real-world recommendation tasks, many deep neural network based Collaborative Filtering (CF) models have been developed to project user-item interactions into latent feature space, based on various neural architectures, such as multi-layer perceptron, autoencoder, and graph neural networks. However, the majority of existing collaborative filtering systems are not well designed to handle missing data. Particularly, in order to inject the negative signals in the training phase, these solutions largely rely on negative sampling from unobserved user-item interactions and simply treating them as negative instances, which brings the recommendation performance degradation. To address the issues, we develop a C ollaborative R eflection-Augmented A utoencoder N etwork (CRANet), that is capable of exploring transferable knowledge from observed and unobserved user-item interactions. The network architecture of CRANet is formed of an integrative structure with a reflective receptor network and an information fusion autoencoder module, which endows our recommendation framework with the ability of encoding implicit user’s pairwise preference on both interacted and non-interacted items. Additionally, a parametric regularization-based tied-weight scheme is designed to perform robust joint training of the two-stage CRANetmodel. We finally experimentally validate CRANeton four diverse benchmark datasets corresponding to two recommendation tasks, to show that debiasing the negative signals of user-item interactions improves the performance as compared to various state-of-the-art recommendation techniques. Our source code is available at https://github.com/akaxlh/CRANet.

A genetic-based pairwise trip planner recommender system

The massive growth of internet users nowadays can be a big opportunity for the businesses to promote their services. This opportunity is not only for e-commerce, but also for other e-services, such as e-tourism. In this paper, we propose an approach of personalized recommender system with pairwise preference elicitation for the e-tourism domain area. We used a combination of Genetic Agorithm with pairwise user preference elicitation approach. The advantages of pairwise preference elicitation method, as opposed to the pointwise method, have been shown in many studies, including to reduce incosistency and confusion of a rating number. We also performed a user evaluation study by inviting 24 participants to examine the proposed system and publish the POIs dataset which contains 201 attractions used in this study.

MUPPscore: An R script for expected a posteriori scoring of multidimensional pairwise preference items

In this manual, we present a flexible and freely available tool for obtaining latent trait scores from multi-unidimensional pairwise preference (MUPP) tests: An R script named MUPPscore. The development of the MUPPscore script provides a solution to the issue that is the previously inconvenient estimation of forced choice item pairs. Instead of using the computationally-intensive multidimensional Bayes modal procedure, the MUPPscore script employs the expected a posterior (EAP) scoring procedure, which provides plausible latent trait score estimates and is also consistent with scoring algorithms used in existing software programs intended for single stimulus measures (e.g., GGUM2004, IRTPRO). The MUPPscore script also returns the empirical marginal reliability of EAP theta estimates and outputs a series of files that can be used to easily create and modify three-dimensional surface charts for plotting MUPP item response function (IRF) in Microsoft Excel.

A Genetic-Based Pairwise trip Planner Recommender System

The massive growth of internet users nowadays can be a big opportunity for the businesses to promote their services. This opportunity is not only for e-commerce, but also for other e-services, such as e-tourism. In this paper, we propose an approach of personalized recommender system with pairwise preference elicitation for the e-tourism domain area. We used a combination of Genetic Agorithm with pairwise user preference elicitation approach. The advantages of pairwise preference elicitation method, as opposed to the pointwise method, have been shown in many studies, including to reduce incosistency and confusion of a rating number. We also performed a user evaluation study by inviting 24 participants to examine the proposed system and publish the POIs dataset which contains 201 attractions used in this study.

Very rapid multi-odour discrimination learning in the ant Lasius niger

Insects can be very good learners. For example, they can form associations between a cue and a reward after only one exposure. Discrimination learning, in which multiple cues are associated with different outcomes, is critical for responding correctly complex environments. However, the extent of such discrimination learning is not well explored. Studies concerning discrimination learning within one valence are also rare. Here we ask whether Lasius niger ants can form multiple concurrent associations to different reward levels, and how rapidly such associations can be learned. We allowed individual workers to sequentially feed on up to four different food qualities, each associated with a different odour cue. Using pairwise preference tests, we found that ants can successfully learn at least two, and likely three, odour/quality associations, requiring as little as one exposure to each combination in order for learning to take place. By testing preference between two non-extreme values (i.e. between 0.4 M and 0.8 M having been trained to the qualities 0.2, 0.4, 0.8 and 1.6) we exclude the possibility that ants are only memorising the best and worst values in a set. Such rapid learning of multiple associations, within one valence and one modality, is impressive, and makes Lasius niger a very tractable model for complex training paradigms.

PrefMap: Visualization of fuzzy pairwise preference structures

Fuzzy pairwise preferences are an important model to specify and process expert opinions. A fuzzy pairwise preference matrix contains degrees of preference of each option over each other option. Such degrees of preference are often numerically specified by domain experts. In decision processes it is highly desirable to be able to analyze such preference structures, in order to answer questions like: Which objects are most or least preferred? Are there clusters of options with similar preference? Are the preferences consistent or partially contradictory? An important approach for such analysis is visualization. The goal is to produce good visualizations of preference matrices in order to better understand the expert opinions, to easily identify favorite or less favorite options, to discuss and address inconsistencies, or to reach consensus in group decision processes. Standard methods for visualization of preferences are matrix visualization and chord diagrams, which are not suitable for larger data sets, and which are not able to visualize clusters or inconsistencies. To overcome this drawback we propose PrefMap, a new method for visualizing preference matrices. Experiments with nine artificial and real–world preference data sets indicate that PrefMap yields good visualizations that allow to easily identify favorite and less favorite options, clusters, and inconsistencies, even for large data sets.

SetRank: A Setwise Bayesian Approach for Collaborative Ranking from Implicit Feedback

The recent development of online recommender systems has a focus on collaborative ranking from implicit feedback, such as user clicks and purchases. Different from explicit ratings, which reflect graded user preferences, the implicit feedback only generates positive and unobserved labels. While considerable efforts have been made in this direction, the well-known pairwise and listwise approaches have still been limited by various challenges. Specifically, for the pairwise approaches, the assumption of independent pairwise preference is not always held in practice. Also, the listwise approaches cannot efficiently accommodate “ties” due to the precondition of the entire list permutation. To this end, in this paper, we propose a novel setwise Bayesian approach for collaborative ranking, namely SetRank, to inherently accommodate the characteristics of implicit feedback in recommender system. Specifically, SetRank aims at maximizing the posterior probability of novel setwise preference comparisons and can be implemented with matrix factorization and neural networks. Meanwhile, we also present the theoretical analysis of SetRank to show that the bound of excess risk can be proportional to √M/N, where M and N are the numbers of items and users, respectively. Finally, extensive experiments on four real-world datasets clearly validate the superiority of SetRank compared with various state-of-the-art baselines.