| 摘要: |
| 历史街区承载着丰富的物质与非物质文化要素，其演化过程呈现出类似生命体的“基因”特征。通过语义网络这一知识表示方式，构建面向历史街区空间的基因模型，目的是通过语义网络解析为历史街区更新提供参考策略。首先通过语义网络工具对东关街历史街区关键基因要素进行全口径交叉识别与提取。根据类型学原理、空间度序列等原理，对街区肌理、空间序列、建筑风格、材料装饰等维度进行解析，区别显性基因和隐性基因，构建街区基因的语义网络模型和模型母版。然后从总体、中观、微观层面，以街区肌理结构、建筑组团布局和建筑立面细节为例，进行肌理基因特征分析、组团空间分析和特征要素拓扑量化分析。最后，再次从总体、中观、微观层面，分别以街区肌理织补、基因原型演绎、基因重塑转译三方面为例，探讨语义网络法在东关街的更新实践中的作用。通过研究，证明语义网络解析方法可以用于历史街区基因的表达和量化，可以兼容显性、隐性基因的形式化分析，为量化分析提供必要的参考。同时，语义网络解析补充了历史街区更新方法，为历史街区更新研究提供了新的视角。 |
| 关键词: 大连东关街 语义网络解析 街区基因 历史街区更新 |
| DOI： |
| 分类号:TU981 |
| 基金项目:国家社科基金一般项目“我国中心城市的韧性城市空间重构体系研究”（21BSH039）） |
|
| Semantic Network Analysis and Renewal Strategies of Historical District Genes ——A Case Study of Dongguan Street in Dalian |
|
Qi Lin,Tang Jian,Dong Jun
|
|
Department of Public and Social Management Research, Party School of the CPC Dalian Municipal Committee, Dalian 116013, China
|
| Abstract: |
| Abstract：Historical districts embody rich tangible and intangible cultural elements, exhibiting "gene"-like characteristics in their evolutionary processes. Academia has developed a relatively mature theory of "spatial genes" for such districts, ranging from typology at the morphological level to intangible cultural genes. Research has also emerged on techniques for gene identification, extraction, and application in renewal practices. However, as a multi-dimensional complex system, the diverse elements within a historical district gain meaning only when integrated into a whole, fundamentally reliant on the relationships between elements. This necessitates applying semantic network analysis to parse the genes of historical districts. This approach not only focuses on genetic elements but also incorporates their interrelationships into the defining characteristics of the district's genes. Semantic networks serve as a tool for knowledge representation. The objective of constructing a spatial gene model for historical districts using semantic networks is to provide reference strategies for their renewal through semantic network analysis.
Firstly, the semantic network tool was used to comprehensively identify and extract key genetic elements of the Dongguan Street historical district via cross-referencing. Relevant literature was retrieved from databases, and ROST CM6 was employed for keyword extraction and frequency statistics. High-frequency words and their co-occurrence relationships were then imported into Gephi to construct a semantic network and identify primary influencing factors. Following element extraction, a semantic network model for the district's genes was built. Spatial entities like streets, courtyards, and buildings extracted from Dongguan Street's urban fabric, alongside concepts from other dimensions such as style, function, and materials, served as nodes in the model. Based on the identified genetic elements, diagrams were created to parse dimensions including urban fabric morphology, spatial sequences, architectural styles, and material decorations. Concurrently, intangible genetic elements were analyzed, combining with tangible ones to construct a characteristic pedigree of Dongguan Street's genetic elements. To mitigate analytical bias arising from superficial differences across dimensions and perspectives, a fundamental semantic network model of district genetic elements was established. During modeling, meso- and micro-level element sets were temporarily folded or ignored, constructing an overall "semantic network master template" by combining prototypical gene fragments within a defined scope. Subsequently, analysis was conducted at the macro, meso and micro levels, exemplified by urban fabric structure, building cluster layout, and architectural fa?ade decoration/components. Macro-level analysis involved comparing Dongguan Street with typical Eastern and Western urban districts to extract its key fabric characteristics. At the meso-level, the No. 18 Courtyard was analyzed as a typical gene fragment, focusing on spatial node degree, degree distribution, and centrality. At the micro-level, decorations and components of buildings in plot 4-3 were subjected to topologically weighted quantitative analysis. The primary fa?ade was analyzed to identify and extract genetic elements, leading to a semantic network model for the fa?ade. An Analytical Hierarchy Process (AHP) was used to establish an indicator system for fa?ade elements and determine their weights. Finally, guided by the Dalian Historic and Cultural City Protection Plan (2020) and the Revised Protection Plan for Dongguan Street Historic and Cultural District (2022), the study explored the supplementary application of this methodology in Dongguan Street's renewal at the macro, meso, and micro levels. Practical application research focused on three main genetic characteristics through texture darning, prototype derivation, and genetic element translation, corresponding to urban fabric structure, building cluster layout, and architectural fa?ade details/components.
In texture darning (urban fabric repair): Sixteen developable texture darning plots and five restorative historic texture darning plots were identified within the district. Block planning and design were implemented to continue historical spatial morphology and maintain factors such as street wall continuity. In prototype derivation (of genetic archetypes): Supported by the semantic network master template, the genetic prototype and implicit genetic prototype fragments of the courtyard cluster's spatial inner layout were derived. Taking metrics such as degree distribution and vertex degree as examples, and integrating other explicit genes, the implicit genes of the block were consolidated. This informed the comprehensive renewal of the No. 18 Courtyard. In genetic element translation: Buildings were categorized into immovable cultural relics, historic buildings, modifiable buildings, and valueless buildings. Corresponding approaches including form replication, moderate adaptation, and demolition with reconstruction were employed to implement micro-level updates to individual structures and architectural details.
The structured quantitative model of Dongguan Street's genetic elements provides a valuable reference for district renewal planning. The study demonstrates that the semantic network method is suitable for expressing and quantifying historical district genes. It offers a formal analytical framework compatible with both tangible and intangible genes, providing essential references for district renewal and introducing a novel research perspective. Furthermore, the inherent semantic expression advantage of semantic networks allows for conversion into machine language, providing a human-machine collaborative interface for this analytical approach. Future research requires further exploration and technological refinement to address more complex challenges. |
| Key words: Dongguan Street Semantic Network Analysis District Genes Historic District Renewal |
