Personalizing Graph Neural Networks with Attention Mechanism for Session-based Recommendation

The problem of personalized session-based recommendation aims to predict users' next click based on their sequential behaviors. Existing session-based recommendation methods only consider all sessions of user as a single sequence, ignoring the relationship of among sessions. Other than that, most of them neglect complex transitions of items and the collaborative relationship between users and items. To this end, we propose a novel method, named Personalizing Graph Neural Networks with Attention Mechanism, A-PGNN for brevity. A-PGNN mainly consists of two components: One is Personalizing Graph Neural Network (PGNN), which is used to capture complex transitions in user session sequence. Compared with the traditional Graph Neural Network (GNN) model, it also considers the role of users in the sequence. The other is Dot-Product Attention mechanism, which draws on the attention mechanism in machine translation to explicitly model the effect of historical sessions on the current session. These two parts make it possible to learn the multi-level transition relationships between items and sessions in user-specific fashion. Extensive experiments conducted on two real-world data sets show that A-PGNN significantly outperforms the state-of-the-art personalizing session-based recommendation methods consistently.

A Situation Calculus-based Approach To Model Ubiquitous Information Services

This paper presents an augmented situation calculus-based approach to model autonomous computing paradigm in ubiquitous information services. To make it practical for commercial development and easier to support autonomous paradigm imposed by ubiquitous information services, we made improvements based on Reiter's standard situation calculus. First we explore the inherent relationship between fluents and evolution: since not all fluents contribute to systems' evolution and some fluents can be derived from some others, we define those fluents that are sufficient and necessary to determine evolutional potential as decisive fluents, and then we prove that their successor states wrt to deterministic complex actions satisfy Markov property. Then, within the calculus framework we build, we introduce validity theory to model the autonomous services with application-specific validity requirements, including: validity fluents to axiomatize validity requirements, heuristic multiple alternative service choices ranging from complete acceptance, partial acceptance, to complete rejection, and validity-ensured policy to comprise such alternative service choices into organic, autonomously-computable services. Our approach is demonstrated by a ubiquitous calendaring service, ACS, throughout the paper.