With the intelligent transformation of communication networks, current risk management systems for telecom operators face challenges such as the mismatch between static risk assessment models and dynamic business logic evolution, as well as the inability to accurately and dynamically perceive multisource cross-domain risks. To address these issues, this paper proposes a multi-perspective spatiotemporal graph neural network model based on knowledge graphs. First, a generative adversarial network (GAN) is employed to reconstruct and impute missing values in key operational features. Next, a knowledge graph is utilized to model the correlations among multi-source influencing factors, generating a deep relational graph of project dependencies. Finally, spatial graph convolution and attention mechanisms are applied to capture deep risk information from multiple semantic perspectives, while temporal representation methods are used to extract time-dependent correlations in project risk data, thereby predicting risks for all projects. Experiments on real-world datasets from Hubei Telecom's enterprise customer service operations demonstrate that the proposed method achieves high accuracy and enables precise perception of multi-source cross-domain risks.

Modern Industrial Control Systems (ICS) have become prime targets for cyberattacks due to the increasing connectivity and complexity of operational infrastructures. These attacks can have severe consequences, making intrusion detection crucial. Recent deep learning approaches have steadily improved intrusion detection accuracy. However, they often fail to model the relationships between multiple sensors and actuators, lack the integration of physical priors, and demonstrate limited adaptability to varying conditions. In this paper, we propose GECAT (Graph-Enhanced Causality-Aware Transformer), a novel intrusion detection framework tailored for ICS data streams. Our approach combines graph-based sensor relationship modeling, causal inference principles, and a specially designed temporal transformer backbone to robustly detect and classify complex attacks.We integrate physical domain knowledge with data-driven graph learning modules to create an adaptive representation that adapts to diverse scenarios and enforces realistic process constraints. Experimental results on SWaT and WADI demonstrate that GECAT significantly outperforms a comprehensive set of state-of-the-art baselines across multiple metrics, indicating its effectiveness in defending ICS against sophisticated adversaries. Additionally, extensive ablation studies confirm the distinct contributions of each core component in our design.

Hybrid Transactional and Analytical Processing (HTAP) databases require processing transactional and analytical queries in isolation to remove the interference between them. To achieve this, it is necessary to maintain different replicas of data specified for the two types of queries. However, it is challenging to provide a consistent view for distributed replicas within a storage system, where analytical requests can efficiently read consistent and fresh data from transactional workloads at scale and with high availability. To meet this challenge, we propose extending replicated state machine-based consensus algorithms to provide consistent replicas for HTAP workloads. Based on this novel idea, we present a Raft-based HTAP database: TiDB. In the database, we design a multi-Raft storage system which consists of a row store and a column store. The row store is built based on the Raft algorithm. It is scalable to materialize updates from transactional requests with high availability. In particular, it asynchronously replicates Raft logs to learners which transform row format to column format for tuples, forming a real-time updatable column store. This column store allows analytical queries to efficiently read fresh and consistent data with strong isolation from transactions on the row store. Based on this storage system, we build an SQL engine to process large-scale distributed transactions and expensive analytical queries. The SQL engine optimally accesses row-format and column-format replicas of data. We also include a powerful analysis engine, TiSpark, to help TiDB connect to the Hadoop ecosystem. Comprehensive experiments show that TiDB achieves isolated high performance under CH-benCHmark, a benchmark focusing on HTAP workloads.

The task of machine reading comprehension (MRC) has evolved from answering simple questions from well-edited text to answering real questions from users out of web data.In the real-world setting, full-body text from multiple relevant documents in the top search results are provided as context for questions from user queries, including not only questions with a single, short, and factual answer, but also questions about reasons, procedures, and opinions.In this case, multiple answers could be equally valid for a single question and each answer may occur multiple times in the context, which should be taken into consideration when we build MRC system.We propose a multi-answer multi-task framework, in which different loss functions are used for multiple reference answers.Minimum Risk Training is applied to solve the multi-occurrence problem of a single answer.Combined with a simple heuristic passage extraction strategy for overlong documents, our model increases the ROUGE-L score on the DuReader dataset from 44.18, the previous state-of-the-art, to 51.09.

Aerosol optical thickness retrieval (AOT) from remotely sensed data over land is still a difficult task because the solar light reflected by the Earth-atmospheric system mainly comes from the ground surface. A novel aerosol remote sensing model SYNTAM by exploiting the synergy of TERRA and AQUA MODIS data could be used to accomplish part of the task of aerosol retrieval over land, especially over higher reflective surface. This paper addressed the issue how the SYNTAM method is applied to the national wide MODIS data for retrieval of aerosol optical thickness over China. The retrieval process is time consuming and the EMS (enhanced memory systems) memory required is very high. To solve the problem, we adapted the SYTAM model to the Grid computing environment. First retrieval process is laid out. The process is decomposed into small steps in pipeline. Decomposing strategy does not consider the difference computability among the computing elements. A coordinator is in charge of load balance. And then the detailed data query, data pre-processing, job monitoring and post-processing are discussed. We have developed one middleware by which the retrieval algorithm can be parallel processed. Preliminary experiment of aerosol retrieval of nationwide range was conducted. The experimental results indicated the computational process had been accelerated in an effective way. Using Grid computing platform, wide range and real-time aerosol information can be retrieved.