Latest ArticlesIn current era, marked by advancements and achievements made in digital and intelligent fields, artificial intelligence (AI) has emerged as a pivotal engine driving technological innovation, which indicates encapsulating and examining the latest trends and future trajectories in AI research makes sense on the development of future AI research. This can be implemented by collecting the research outcomes during recent three years from top-tier international conferences and journals in the field of AI that are recommended by the China Computer Federation (CCF-A category), introducing keyword-centric analyses based on a bibliometric methodologies, and analyzing research hotspots based on high-frequency keywords, discerning emerging trends through newly-added keywords, identifying high-impact studies using citation-weighted keyword analysis. The result of these analyses, which contains significant information about trends in AI research, can enable the principal directions of AI research to be delineated and the interconnections and integrative fusion within mainstream AI research directions to be unveiled. Moreover, an in-depth exploration of the current hot topics, such as large language models (LLMs), AI-driven scientific research (AI for Science) and visual generation technologies, would help us reveal the underlying scientific theories and application prospects behind these technological innovations, thereby the latest trends and future trajectories in AI field get demonstrated more adequately and concretely.
Accurately quantifying the relationship between skills and salary is essential to improve reasonable job salary setting and promote talent attraction and retention. However, the relationship between skills and salary is complex because it involves modeling set utility in a high-dimensional space with massive possible elements. Deep neural networks offer a new solution for complex fitting problems. However, for skill-based fine-grained salary prediction, there still lacks interpretable neural networks that can effectively model set utility under the influence of complex variables. To address this issue, we propose a marginal contribution-based incremental set utility network (MCISUN). MCISUN models the marginal contribution of elements when they are added to the set. In this way, the set utility can be naturally obtained in a flexible and interpretable way. In particular, rather than relying on pooling structures to ensure permutation invariance, MCISUN constructs order-sensitive intermediate results through recurrent attention neural networks and takes advantage of the sets’ permutation invariance property to achieve data augmentation, thus improving the model’s robustness. We conduct extensive experiments on a real-world large-scale salary dataset. The experimental results show that MCISUN outperforms state-of-the-art models by 30% for skill-based salary prediction. Qualitative experiments show that our model can recognize reasonable skill contribution values and capture the relationship between skills.
Learning effective high-order feature interactions is crucial for click through rate (CTR) prediction in recommender systems. Existing methods that learn meaningful high-order feature combinations by reassembling low-order feature combinations, i.e., 2-order feature interaction, suffer from high computational costs to calculate the interaction weight of all pairwise feature interactions. Some deep neural network-based methods can be seen as universal function approximators to potentially learn all kinds of feature interactions. However, it had been proved to be inefficient to approximate the low-order interactions, i.e., 2-order or 3rd-order feature interactions, which may influence the accuracy of CTR prediction task. Based on the above consideration, we propose a multi-granularity based feature interaction pruning network (FeatNet) for CTR prediction task. Firstly, FeatNet generates different subsets with a threshold pruning operation to select the meaningful feature combinations on the explicit feature granularity, which enables FeatNet to keep the diversity of different feature combinations, and reduce the complexity of high-order feature interactions. Based on the pruned feature subsets, implicit high-order feature interactions are further conducted on the granularity of feature elements, which automatically filters out the invalid feature interactions. Extensive experiments are conducted on two real-world datasets, showing the superiority of FeatNet in CTR prediction.
Conversational emotion recognition is the task of classifying emotions based on conversations. The conversation data are characterized by colloquial language and a wide range of topics, with semantic similarities among labels. Colloquial language exhibits issues such as word ambiguity and the omission of semantic information, emphasizing the importance of common sense and grammatical knowledge in conversational emotion recognition tasks, and these factors enable the model to accurately capture semantic information. Moreover, the current challenge lies in the variations in text richness and the frequency of emotion transfer across different dialogue scenarios, which result in suboptimal classification performance. We propose CK-ERC model to address these challenges. In the pre-training phase, CK-ERC model extracts structured data to incorporate common sense knowledge graphs and grammatical knowledge graphs, aiding the model in accurately capturing colloquial information. In the fine-tuning phase, a supervised contrast learning task is introduced to help the model identify similar emotional labels. Furthermore, a dynamic threshold-based curriculum learning strategy is designed for training and optimizing the model based on text richness (from high to low) and emotion transfer frequency (from low to high). CK-ERC model demonstrates superior performance in various conversation modes, including two-person conversation, multi-person conversation, simulated conversation, and daily conversation. Particularly, CK-ERC model achieves the best performance on MELD and EmoryNLP datasets.
Multimodal machine learning represents a novel paradigm in artificial intelligence, leveraging various modalities and intelligent processing algorithms to achieve enhanced performance. Multimodal representation and fusion are two pivotal tasks in multimodal machine learning. Currently, most multimodal representation methods pay little attention to inter-sample collaboration, leading to a lack of robustness in feature representation. Additionally, most multimodal feature fusion methods exhibit sensitivity to noisy data. Therefore, in the realm of multimodal representation, an approach based on both intra-sample and inter-sample multimodal collaboration is proposed to facilitate a comprehensive understanding of interactions within and between modalities, ultimately enhancing the robustness of feature representation. Firstly, text, speech, and visual features are individually extracted based on pre-trained models such as BERT, Wav2vec 2.0, and Faster R-CNN. Subsequently, considering the complementarity and consistency of multimodal data, two categories of encoders, namely modality-specific and modality-shared, are constructed to learn both modality-specific and shared feature representations. Furthermore, intra-sample collaboration loss functions are formulated using central moment differences and orthogonality, while inter-sample collaboration loss functions are established using contrastive learning. Lastly, a representation learning function is designed based on intra-sample collaboration, inter-sample collaboration, and sample reconstruction errors. Regarding multimodal fusion, an adaptive multimodal feature fusion method is designed, accounting for the possibility that each modality may exhibit varying types of effects and levels of noise at different times, using attention mechanisms and gated neural networks. Experimental results on the multimodal intent recognition dataset MIntRec and emotion datasets CMU-MOSI and CMU-MOSEI demonstrate that this multimodal learning approach outperforms baseline methods across multiple evaluation metrics.
Police dossier, as one of the main records handled by the police department, contains massive and crucial policing information. As an important means, efficient information extraction from police dossier is helpful for case analysis, crime trend prediction, and the improvement of the public security management. However, the text of police dossier is written by police officers using natural language, which makes it difficult to extract crucial information. Traditional information extraction of police dossier heavily relies on manual effort and predefined templates, resulting in low efficiency and poor generality. Considering the particularity of police dossier, in this paper, a knowledge extraction method based on deep semantics analysis is proposed. This method consists of two core tasks: named entity recognition and relation extraction. Focusing on Chinese text, we propose a named entity recognition method that integrates structural and glyph features of Chinese characters. On the basis of entity recognition results, with the help of a specially constructed policing thesaurus, a relationship extraction method combining rule based and trigger word is proposed. Both on a publicly available Weibo dataset and a real dossier dataset provided by our partner a local police department, compared with several baseline named entity recognition models, our proposed method shows better performance in classifying exact entities and finding more potential entities. The automatically extracted relationships have also been verified and committed by the police department branch. A particular information system has been used in practice.
In this paper, the mutually beneficial relationship between depth estimation and semantic segmentation is investigated, and a self-supervised monocular depth estimation method for joint semantic segmentation USegDepth is proposed. The shared encoder for semantic segmentation and depth estimation is implemented to achieve semantic guidance. To further improve the across multiple tasks performance of the encoder, a multi-task feature extraction module is designed. The module is stacked to generate the shared encoder, solving the poor feature representation problem of the model due to limited receptive field and lack of cross-channel interaction, and the model accuracy is improved further. And a cross-task interaction module is proposed for bidirectional cross-domain information interaction to refine the depth features, improving depth estimation performance, especially in weak texture regions and object boundaries with limited luminosity consistency supervision. Through training and evaluation on KITTI dataset, the experimental results show that the mean square relative error of USegDepth is reduced by 0.176 percentage points compared with that of SGDepth, and the threshold accuracy reaches 98.4% at a threshold value of 1.253, proving the high accuracy of USegDepth in depth prediction.
The large models represented by ChatGPT have attracted a lot of attention from industry and academia for their excellent performance on text generation and semantic understanding tasks. The number of large model parameters has increased tens of thousands of times in three years and is still growing, which brings new challenges to storage systems. First, we analyze the storage challenges of large model training, pointing out that large model training has unique computation patterns, storage access patterns, and data characteristics, which makes traditional storage techniques inefficient in handling large model training tasks. Then, we describe three types of storage acceleration techniques and two types of fault-tolerant techniques. The storage acceleration techniques for large model training include: 1) distributed storage technique based on large model computation patterns designs the partitioning, storage, and transferring strategies of model data in distributed clusters based on the partitioning of large model computation tasks and the dependencies between computation tasks; 2) heterogeneous storage access pattern-aware technique for large model training develops data prefetching and transferring strategies among heterogeneous devices with the predictability of storage access patterns in large model training; 3) large model data reduction technique reduces the data size in the model training process according to the characteristics of large model data. The storage fault-tolerant techniques for large model training include: 1) parameter checkpointing technique stores the large model parameters to persistent storage devices; 2) redundant computation technique computes the same version of parameters repeatedly in multiple GPUs. Finally, we give the summary and suggestions for future research.
Graph application is an important branch in the field of big data. Although graph analysis has more significant performance advantages than traditional relational databases in displaying the relationship between entities, the irregular memory access pattern caused by a large number of random accesses in graph processing destroys the time and space locality of memory access, thus causing great performance pressure on the off-chip memory system. Therefore, how to correctly measure the performance of graph application in memory system is crucial for efficient graph application architecture optimization. As an extension of average memory access time (AMAT), concurrent average memory access time (C-AMAT) takes into account the locality and concurrency of memory access, and can more accurately evaluate and analyze the performance of modern processors in the storage system. However, the C-AMAT model ignores the fact that the lower-level cache layer of the processor accesses serially, which will lead to the inaccuracy of the calculation. At the same time, it is difficult to obtain the parameters required for the calculation due to the “pure miss cycle” and other reasons, which also makes it difficult for C-AMAT to be applied in practice. In order to match the computing model of C-AMAT with the memory access mode in modern computers, we propose parallel C-AMAT (PC-AMAT) and serial C-AMAT (SC-AMAT) based on C-AMAT. PC-AMAT and SC-AMAT respectively extend and characterize the computing model of C-AMAT from the parallel and serial access modes of cache. On this basis, we design and implement a “pure miss cycle” extraction algorithm to avoid the huge hardware overhead caused by direct measurement. The experimental results show that the correlation between PC-AMAT and SC-AMAT, and IPC is stronger than that of C-AMAT in single-core and multi-core mode. Finally, PC-AMAT and SC-AMAT are used to measure and analyze the memory performance of graph application, based on which the optimization strategy of graph application access is proposed.
With the development of AI democratization, deep neural networks (DNNs) have been widely applied to edge devices, such as smart phones and automated driving, etc. Stochastic computing (SC) as a promising technique performs fundamental machine learning (ML) tasks using simple logic gates instead of complicated binary arithmetic circuits. SC has advantages of low-power and low-cost DNNs execution on edge devices with constrained resources (e.g., energy, computation and memory units, etc.). However, previous SC work only designs one group of setting for fixed hardware implementation, ignoring the dynamic hardware resources (e.g., battery), which leads to low hardware efficiency and short battery life. In order to save energy for battery-powered edge devices, dynamic voltage and frequency scaling (DVFS) technique is widely used for hardware reconfiguration to prolong battery life. In this paper, we creatively propose a run-time reconfigurable framework, namely RR-SC, for SC-based DNNs and first attempt to combine hardware and software reconfigurations to satisfy the time constraint of inference and maximally save energy. RR-SC using reinforcement learning (RL) can generate multiple groups of model settings at one time, which can satisfy the accuracy constraints under different hardware settings (i.e., different voltage/frequency levels). The solution has the best accuracy and hardware efficiency trade-off. Meanwhile, the model settings are switched on a backbone model at run-time, which enables lightweight software reconfiguration. Experimental results show that RR-SC can switch the lightweight settings within 110 ms to guarantee the required real-time constraint at different hardware levels. Meanwhile, it can achieve up to 7.6 times improvement for the number of model inference with only 1% accuracy loss.