Recent advancements have highlighted that Large Language Models (LLMs) are prone to hallucinations when solving complex reasoning problems, leading to erroneous results. To tackle this issue, researchers incorporate Knowledge Graphs (KGs) to improve the reasoning ability of LLMs. However, existing methods face two limitations: 1) they typically assume that all answers to the questions are contained in KGs, neglecting the incompleteness issue of KGs, and 2) they treat the KG as a static repository and overlook the implicit logical reasoning structures inherent in KGs. In this paper, we introduce SymAgent, an innovative neural-symbolic agent framework that achieves collaborative augmentation between KGs and LLMs. We conceptualize KGs as dynamic environments and transform complex reasoning tasks into a multi-step interactive process, enabling KGs to participate deeply in the reasoning process. SymAgent consists of two modules: Agent-Planner and Agent-Executor. The Agent-Planner leverages LLM's inductive reasoning capability to extract symbolic rules from KGs, guiding efficient question decomposition. The Agent-Executor autonomously invokes predefined action tools to integrate information from KGs and external documents, addressing the issues of KG incompleteness. Furthermore, we design a self-learning framework comprising online exploration and offline iterative policy updating phases, enabling the agent to automatically synthesize reasoning trajectories and improve performance. Experimental results demonstrate that SymAgent with weak LLM backbones (i.e., 7B series) yields better or comparable performance compared to various strong baselines. Further analysis reveals that our agent can identify missing triples, facilitating automatic KG updates.

Aggregating distributed energy resources in power systems significantly increases uncertainties, in particular caused by the fluctuation of renewable energy generation. This issue has driven the necessity of widely exploiting advanced predictive control techniques under uncertainty to ensure long-term economics and decarbonization. In this paper, we propose a real-time uncertainty-aware energy dispatch framework, which is composed of two key elements: (i) A hybrid forecast-and-optimize sequential task, integrating deep learning-based forecasting and stochastic optimization, where these two stages are connected by the uncertainty estimation at multiple temporal resolutions; (ii) An efficient online data augmentation scheme, jointly involving model pre-training and online fine-tuning stages. In this way, the proposed framework is capable to rapidly adapt to the real-time data distribution, as well as to target on uncertainties caused by data drift, model discrepancy and environment perturbations in the control process, and finally to realize an optimal and robust dispatch solution. The proposed framework won the championship in CityLearn Challenge 2022, which provided an influential opportunity to investigate the potential of AI application in the energy domain. In addition, comprehensive experiments are conducted to interpret its effectiveness in the real-life scenario of smart building energy management.

The recommender system for e-commerce aims to recommend appropriate items to online customers in order to drive more views, clicks or purchases on those items. Most existing models incorporate the users' historical behaviors, their profiles, and the item metadata to achieve good performances. However, since more and more people are surfing the Internet without logging in, it is no longer capable to provide accurate recommendations based on the historical data or profiles. To tackle this issue, we propose MentalNet-a mental model for e-commerce recommendation by enhancing the gated Graph Neural Network (GNN) and capturing user intent in a short session. More precisely, MentalNet is composed of two stages: in the first stage, we enhance the gated GNN to take into account the complex graph-level transitions among items, for an improved item representation; In the second stage, we propose a mental model to simulate user intent using item embedding, and then compute item preferences based on each intent. Finally, we empirically demonstrate the effectiveness of the proposed method on three datasets, including the CIKM CUP data, the RecSys Challenge data and a real-world e-commerce dataset in Alibaba Group.

Sequential prediction is one of the key components in recommendation. In online e-commerce recommendation system, user behavior consists of the sequential visiting logs and item behavior contains the interacted user list in order. Most of the existing state-of-the-art sequential prediction methods only consider the user behavior while ignoring the item behavior. In addition, we find that user behavior varies greatly at different time, and most existing models fail to characterize the rich temporal information. To address the above problems, we propose a transformer-based spatial-temporal recommendation framework (STEM). In the STEM framework, we first utilize attention mechanisms to model user behavior and item behavior, and then exploit spatial and temporal information through a transformer-based model. The STEM framework, as a plug-in, is able to be incorporated into many neural network-based sequential recommendation methods to improve performance. We conduct extensive experiments on three real-world Amazon datasets. The results demonstrate the effectiveness of our proposed framework.

Click-through rate (CTR) prediction is essential in the modelling of a recommender system. Previous studies mainly focus on user behavior modelling, while few of them consider candidate item representations. This makes the models strongly dependent on user representations, and less effective when user behavior is sparse. Furthermore, most existing works regard the candidate item as one fixed embedding and ignore the multi-representational characteristics of the item. To handle the above issues, we propose a Deep multi-Representational Item NetworK (DRINK) for CTR prediction. Specifically, to tackle the sparse user behavior problem, we construct a sequence of interacting users and timestamps to represent the candidate item; to dynamically capture the characteristics of the item, we propose a transformer-based multi-representational item network consisting of a multi-CLS representation submodule and contextualized global item representation submodule. In addition, we propose to decouple the time information and item behavior to avoid information overwhelming. Outputs of the above components are concatenated and fed into a MLP layer to fit the CTR. We conduct extensive experiments on real-world datasets of Amazon and the results demonstrate the effectiveness of the proposed model.