随着隐私保护意识的日益增强以及数据合规要求的严格化，隐私计算作为一种能在数据不离开原始所有者控制的情况下实现数据分析的技术，正逐渐成为解决数据共享与协作中隐私问题的关键手段。然而，随着应用场景的多元化复杂化和用户需求的多样化，隐私计算技术面临着提升易用性、扩大适用性及优化客户体验的轻量化需求。尤其是在资源有限的前端环境和数据处理的后端环境中，如何实现高效、低资源消耗的隐私计算，成为亟待解决的关键问题。本文针对这一需求，深入研究了基于WASM（WebAssembly）和DSL（领域特定语言）的轻量化隐私计算技术，以提升隐私计算的实用性和用户体验，并为构建全面的轻量化隐私计算平台系统提供理论基础和技术方案。首先，前端轻量化隐私计算技术的研究主要聚焦于WASM的应用。WASM作为一种高效、安全且跨平台的二进制指令格式，能够在现代浏览器中近乎原生地运行，为前端环境下的隐私计算提供了理想的执行环境。本文探讨如何将主流隐私计算算法封装为WASM模块，实现浏览器内直接执行，无需依赖复杂插件，从而确保前端应用中数据隐私的实时保护。此外，本文还关注如何通过WASM实现隐私计算任务的安全处理与后台运行，设计简洁易用的JavaScript接口，方便前端开发者集成与调用隐私计算功能。在后端轻量化隐私计算技术方面，本文着重研究基于DSL的隐私计算引擎设计与实现。DSL作为一种专为特定领域或任务设计的语言，能够简化隐私计算算法的表述与实现过程，提高开发效率。本文设计了一种针对隐私计算任务的DSL，其语法简洁直观，能够清晰表述同态加密、隐私信息检索等核心逻辑，有利于编译器进行优化处理。研究深入到DSL编译与执行层面，探讨如何将其编译为高效的目标代码，在后端环境中高效执行，并研究编译优化技术以提升隐私计算任务的执行效率。最后，本文提出了构建轻量化隐私计算平台系统的总体构想。轻量化隐私计算平台系统巧妙地融合了同态加密算法、隐私信息检索技术以及可信执行环境，形成了一套综合全面、灵活适应的隐私计算解决方案，能够有效应对各类需求下的隐私保护与数据处理任务。轻量化隐私计算平台集成了前端轻量化隐私计算技术和后端轻量化隐私计算技术，以实现高效、安全且资源占用较小的轻量化隐私计算技术系统。通过对轻量化隐私计算平台在金融场景下的组合管理应用和隐私信息检索应用，进一步证明了本文技术可用性。

With the increasing awareness of privacy protection and the strictness of data compliance requirements, privacy computing is emerging as a crucial means to address privacy issues in data sharing and collaboration. Because privacy computing enables data analysis without the data leaving the control of its original owner. However, as application scenarios become more diverse and complex and user demands more varied, privacy computing technologies face the need to enhance usability, expand applicability, and optimize customer experience through lightweight solutions. Particularly in resource-constrained front-end environments and data processing back-end environments, achieving efficient, low-resource-consuming privacy computing has become a critical issue that needs urgent resolution. This paper delves into lightweight privacy computing technologies based on WebAssembly （WASM） and Domain-Specific Languages （DSL） to improve the practicality and user experience of privacy computing and provide theoretical foundations and technical solutions for building comprehensive lightweight privacy computing platforms.Firstly, the research on front-end lightweight privacy computing primarily focuses on the application of WASM. WASM, as an efficient, secure, and cross-platform binary instruction format, can run almost natively in modern browsers, providing an ideal execution environment for privacy computing in front-end settings. This paper explores how to encapsulate mainstream privacy computing algorithms into WASM modules for direct execution within the browser, eliminating the need for complex plugins and ensuring real-time data privacy protection in front-end applications. Additionally, the paper addresses how to implement asynchronous processing and background execution of privacy computing tasks through WASM to optimize the front-end interactive experience, and designs simple and user-friendly JavaScript interfaces for front-end developers to easily integrate and invoke privacy computing functionalities.Then, in terms of back-end lightweight privacy computing, the paper focuses on the design and implementation of a privacy computing engine based on DSL. DSL, being a language specifically designed for particular domains or tasks, simplifies the representation and implementation of privacy computing algorithms, thereby improving development efficiency. This paper designs a DSL tailored for privacy computing tasks, featuring concise and intuitive syntax that clearly expresses core logic such as homomorphic encryption and privacy information retrieval, facilitating compiler optimization and parallel processing. The research delves into the compilation and execution layers of DSL, exploring how to compile it into efficient target codefor effective execution in back-end environments, and investigates compilation optimization techniques to enhance the execution efficiency of privacy computing tasks.Last but not least, the paper proposes an overarching vision for building a lightweight privacy computing platform system. This system ingeniously integrates homomorphic encryption algorithms, privacy information retrieval technologies, and trusted execution environments to form a comprehensive, flexible privacy computing solution capable of effectively addressing privacy protection and data processing tasks under various requirements. The lightweight privacy computing platform incorporates both front-end and back-end lightweight privacy computing technologies to achieve an efficient, secure, and low-resource-consuming privacy computing system. The application of the lightweight privacy computing platform in financial scenarios, such as portfolio management systems and privacy information retrieval applications, further demonstrates the feasibility and usability of the technologies discussed in this paper.