Data sharing and exchange play a critical role in promoting the intelligent and digital transformation of government services. However, existing government data sharing and exchange systems typically adopt cascaded architectures, resulting in long supply-demand paths at the architectural level. In addition, limited capabilities in data quality inspection and rapid construction of underlying routing mechanisms lead to low real-time performance, difficulty in ensuring data quality, and inadequate support for scenarios involving large volumes of frequently used data.

To address these challenges, a systematic research approach is adopted. First, guided by the principles of distribution, high reliability, and flexible configuration, a distributed architecture for government data sharing and exchange is proposed. A distributed data exchange network composed of peer nodes is constructed, in which node relationships are equal, thereby shortening data forwarding paths and improving exchange efficiency. By decoupling the control layer from the transport layer, the control layer is dedicated to routing management and node status monitoring, while the transport layer focuses on efficient and reliable data transmission, clarifying the functional structure and operational mechanism of the architecture. Second, a data quality inspection algorithm based on large-model intelligent agents is introduced. Using a unified inspection strategy, the algorithm evaluates data quality across four dimensions—semantic consistency, format standardization, logical consistency, and data integrity—ensuring high-quality data provision. Third, a geographic-aware routing algorithm is proposed by integrating administrative geographic information into distributed Hash tables. A hybrid routing strategy is designed, combining cross-layer routing based on a multiway tree structure with intralayer routing based on a binary tree structure, thereby reducing routing hops during data addressing. Finally, a series of experimental validation processes is employed to verify the effectiveness of key algorithms and the overall architecture.

Compared with the benchmark method, the proposed geographic-aware routing algorithm reduced the average hop count by 76.82%. The intelligent-agent-based data quality inspection algorithm achieved an average precision of 93.06%, an average recall of 93.50%, and an average F₁-score of 0.72. Based on the distributed government data sharing and exchange architecture, three typical business scenarios—real-time transactions, unstructured transactions, and batch transactions—were evaluated. Deployment in Heilongjiang Province enabled on-site performance testing under real-world conditions. The results showed that the average response time for real-time transaction scenarios was 722 ms, with a median of 594 ms. In unstructured transaction scenarios, large-file upload speeds reached 220.0-235.0 MB/s, while batch transaction scenarios achieved an average throughput of 1.5 MB/s and an average write speed of 1961 records/s. Compared with the theoretical performance peak of traditional cascaded systems, the performance was improved by 50% and 91%, respectively.

The proposed distributed government data sharing and exchange system significantly enhances real-time performance and ensures data quality in government data sharing and exchange. It provides a new technical pathway for intelligent and digital government transformation and offers valuable insights into the large-scale application of "artificial intelligence + data elements" in the public sector.

Since the turn of the 21st century, artificial intelligence (AI) has advanced considerably in many domains, including government affairs. Furthermore, the emergence of deep learning has taken the development of many AI fields, including natural language processing (NLP), to a new level. Language models (LMs) are key research directions of NLP. Referred to as statistical models, LMs were initially used to calculate the probability of a sentence; however, in recent years, there have been substantial developments in large language models (LLMs). Notably, LLM products, such as the generative pretrained transformer (GPT) series, have driven the rapid revolution of large language research. Domestic enterprises have also researched LLMs, for example, Huawei's Pangu and Baidu's enhanced language representation with informative entities (ERNIE) bot. These models have been widely used in language translation, abstract construction, named-entity recognition, text classification, and relationship extraction, among other applications, and in government affairs, finance, biomedicine, and other domains.

In this study, we observe that improving the efficiency of governance has become one of the core tasks of the government in the era of big data. With the continuous accumulation of government data, traditional statistical models relying on expert experience and local features gradually suffer limitations during application. However, LLMs, which offer the advantages of high flexibility, strong representation ability, and effective results, can rapidly enhance the intelligence level of government services. First, we review the research progress on early LMs, such as statistical LMs and neural network LMs. Subsequently, we focus on the research progress on LLMs, namely the Transformers series, GPT series, and bidirectional encoder representations from transformers (BERT) series. Finally, we introduce the application of LLMs in government affairs, including government text classification, relationship extraction, public opinion risk identification, named-entity recognition, and government question answering. Moreover, we propose that research on LLMs for government affairs must focus on multimodality, correctly benefit from the trend of "model as a service, " focus on high data security, and clarify government responsibility boundaries. Additionally, a technical path for studying LLMs for government affairs has been proposed.

The application of LLMs in government affairs mainly focuses on small-scale models, lacking examples of application in large-scale models. Compared with smaller models, large models offer many advantages, including high efficiency, broader application scenarios, and more convenience. These advantages can be understood as follows. In terms of efficiency, large models are usually trained on a large amount of heterogeneous data, thus delivering better performance. In terms of application scenarios, large models gradually support multimodal data, resulting in more diverse application scenarios. In terms of convenience, we emphasize the "pretraining + fine-tuning" mode and the invocation method of interfaces, making LLMs more convenient for research and practical applications. This study also analyzes the issues suffered by LLMs, specifically from the technological and ethical perspectives, which have resulted in a panic to a certain extent. For example, ChatGPT has generated many controversies, including whether the generated files are novel, whether using ChatGPT will lead to plagiarism and ambiguity as to who are property rights owners for the generated files. Overall, it can be said that LLMs are in the stage of vigorous development. As the country promotes research on AI and its application in government affairs, LLMs will play an increasingly crucial role in the field.