在我国城镇化高速建设的新阶段，建设过程中特别是建设早期的决策科学化显得愈发重要。然而，长期以来，对于建筑创作合理性这一建筑学基本命题的理论研究尚不够充分，导致本该负责拷问设计依据与设计逻辑的建筑师，事实上却相对被动地接受业主编制的建设项目任务书（也称“设计大纲”，本论文简称“设计任务书”或“任务书”），甚至在错误的题目上开展建筑设计创作；由于这样所产生的设计偏差，极有可能对社会物质环境和国民经济形成危害。 建筑策划是建筑学领域内专门研究如何科学地编制任务书的一套理论和方法。设计任务书的研究是建筑学的本源问题，而设计任务书的编制又是建筑师的基本任务与职责；对设计任务书的评价进行研究，是建筑策划研究的一个发展方向，也是建筑策划评价的核心内容。因此，本研究就这一核心，以任务书为具体研究对象，尝试建立一套评价的科学模式，以充实建筑策划的研究，延伸建筑学的理论框架和设计方法论，拓展建筑师的业务范畴。 本论文采用纵向推进、横向展开的逻辑结构，从理论、法规、实践三个层面进行论述。在理论层面上对建筑策划的经典理论、建筑学领域内外的相关概念和方法进行了梳理，指出建筑设计概念要素已经快速膨胀更新，现有建筑策划技术手段囿于传统数据形式，需要进一步明确建筑策划评价的研究对象，并寻求新的方法和技术刺激其发展。在法规层面，通过对比国内外基本建设流程，指出我国缺少建筑策划及其评审这一必要环节，交叉印证任务书评价研究的必要性和紧迫性，以及制定评价准则的行业需求。在实践层面，通过审视任务书的特征和建筑师自发的任务书评价活动，指出现实中的评价机制过于发散，效率较为低下，急需建立程式化的、具有操作性的任务书评价体系。 在上述三个层面的指导下，本研究收集112份真实的任务书作为样本，参考系统评价学的流程，借鉴风险评估的方法，使用文本数据处理和分析技术，与经典的建筑策划理论相结合，在任务书样本库的数据基础上挖掘评价指标，生成指标权重，建立任务书的评价体系；并面向建筑设计实践，开发指导手册等评价工具。通过在两个建设项目上进行实验和试验性质的应用，验证了本研究所提出的任务书评价体系及应用工具是切实有效的。
In the new stage of rapid urbanization in China, the scientific decision-making in the construction process becomes more and more important, especially in the early stage of design. However, the lack of research on the rationality of architecture design has led to an improper phenomenon in professional practice. The architects who are supposed to be responsible for the validity of design basis and design logic, on the contrary, passively accept design briefs offering by clients themselves; some even do their designs based on wrong conditions and requirements. Due to such way of design, it may easily result in irrational design proposals and unavailable built environment, which are undoubtfully harmful to the social material environment and the national economy. Architectural programming is a set of theories and methods devoted to preparation of design brief scientifically. The design brief is a basic problem of architecture studies. The preparation of design brief is a basic task of professional architects as well. Thus, the evaluation of briefs is a promising niche of architectural programming research field. It is of the key issue of architectural programming evaluation. This paper works on this key issue, taking design brief as a specific object of study, tries to establish a scientific mode of evaluation. This study aims to supplement the researches of architectural programming, expand the theoretical framework of architecture and design methodology, also redefine the scope of architect's services.This paper adopts the logical structure of vertical promotion and horizontal expansion, and discusses from three aspects: theory, regulation and practice. Classic theories of architectural programming, related concepts and methods in and out of the field of architecture are sorted out in the theoretical level. It pointes out that elements of architectural design have been rapidly expanded and updated. Existing techniques of architectural programming are still confined to the traditional data form. which calls for further clarify of research object and new techniques that fit the object. At the level of regulations, by comparing the domestic and foreign construction process, we find that the necessary stage of programming and its evaluation is missing. It states the necessity and urgency of evaluation of briefs and the needs for developing evaluation criteria. At the practical level, by examining features of design brief and spontaneous brief evaluation activities conducted by architects, it stresses that the evaluation in reality is far too divergent, lacking of efficiency. A stylized, operational evaluation system is needed.We collect 112 briefs of practical projects as samples, introduce the process of systematic evaluation and methods of risk assessment, combine techniques of text data mining with classic architectural programming theories. Making use of analysis results out of brief database, we build evaluation indexes and generate weights to establish a standard evaluation system for design brief. We also develop supplementary evaluation tools for architectural design practices. The brief evaluation system and tools given by this study are proved effective through experimental verification and test application carried out on two projects.
