Two analogous tasks have emerged in multimodal dialogue research: multimodal dialogue response generation and multimodal task-oriented dialogue. Both tasks share the goal of response multi-round, interactive content based on the multimodal dialogue history, but the latter focuses on accomplishing specific objectives which can be viewed as the former fine-tuned. The fine-tuning strategy may cause catastrophic forgetting and overfitting on few well-annotated data. Despite considerable progress in both areas, many existing works rely on retrieval-based approaches and additional auxiliary knowledge bases. To address these issues, we propose a Hierarchical Fusion Framework (HFF) for multimodal dialogue response generation. HFF blends these two tasks to learn a generation model from a data-driven perspective by introducing multi-dataset learning scheme, achieving a balance between generalization and expertise. In this work, multi-dataset learning is cast as a multi-objective optimization problem due to potential conflicts between datasets, necessitating a trade-off based on data distribution during training. Hierarchical fusion is performed sequentially between modalities and datasets, which could efficiently establish clear cross-modal relationships and integrate knowledge from multi-dataset. Specifically, HFF aligns extracted unimodal features (image and text) before fusing them through cross-modal attention and integrates them into multimodal encoder-decoder for generating responses. By optimizing the fusion between corpora from multi-dataset as conflicting objectives to satisfy Pareto optimality, our approach effectively facilitates both multimodal task-oriented and task-unoriented dialogues. Experimental results demonstrate the effectiveness of HFF and its comparable performance with all baselines.

Thanks to Network Function Virtualization (NFV), Internet Service Providers (ISPs) can improve network resource utilization with significantly reduced capital and operational expenditures. To dig deeper into the potential of NFV, an important challenge is the resource allocation problem in NFV (NFV-RA), which can be divided into three stages: VNFs chain composition, VNF forwarding graph embedding, and VNFs scheduling. The key to the NFV-RA problem is to design an effective and coordinated resource allocation algorithm for the three stages. Besides, the NFV-RA problem has been proved to be NP-Hard, and thus most existing approaches focus on heuristic and meta-heuristic algorithms. In this paper, we propose an NFV online coordinated resource allocation framework (OCRA) that completes the three stages simultaneously in a coordinated manner by combining parallel Multi-Agent Deep Reinforcement Learning with novel neural networks and RL training techniques. The extensive experimental results show that compared with the state-of-the-art solutions, OCRA is highly-efficient in terms of time, with up to 50% and 10.8% improvement on resource overhead and acceptance ratio, respectively.

Existing algorithms for I/O Linear Temporal Logic (LTL) model checking usually output a single counterexample for a system which violates the property. However, in real-world applications, such as diagnosis and debugging in software and hardware system designs, people often need to have a set of counterexamples or even all counterexamples. For this purpose, we propose an I/O efficient approach for detecting all accepting cycles, called Detecting All Accepting Cycles (DAAC), where the properties to be verified are in LTL. Different from other algorithms for finding all cycles, DAAC first searches for the accepting strongly connected components (ASCCs), and then finds all accepting cycles of every ASCC, which can avoid searching for a great many paths that are impossible to be extended to accepting cycles. In order to further lower DAAC's I/O complexity and improve its performance, we propose an intersection computation technique and a dynamic path management technique, and exploit a minimal perfect hash function (MPHF). We carry out both complexity and experimental comparisons with the state-of-the-art algorithms including Detect Accepting Cycle (DAC), Maximal Accepting Predecessors (MAP) and Iterative-Deepening Depth-First Search (IDDFS). The comparative results show that our approach is better on the whole in terms of I/O complexity and practical performance, despite the fact that it finds all counterexamples.

Model checking is a powerful approach for the formal verification of hardware and software systems. However, this approach suffers from the state space explosion problem, which limits its application to large-scale systems due to space shortage. To overcome this drawback, one of the most effective solutions is to use external memory algorithms. In this paper, we propose an I/O efficient model checking algorithm for large-scale systems. To lower I/O complexity and improve time efficiency, we combine three new techniques: 1) a linear hash-sorting technique; 2) a cached duplicate detection technique; and 3) a dynamic path management technique. We show that the new algorithm has a lower I/O complexity than state-of-the-art I/O efficient model checking algorithms, including detect accepting cycle, maximal accepting predecessors, and iterative-deepening depth-first search. In addition, the experiments show that our algorithm obviously outperforms these three algorithms on the selected representative benchmarks in terms of performance.