范畴符号学
Categorical semiotics
摘要：
从由领域专家描述的不同模型或从由机器学习算法生成的模型中提取的知识的集成受到缺乏适当的框架来指定和集成结构、学习过程、数据转换和数据模型或数据规则的强烈制约。在这项工作中，我们扩展了代数规范方法以用于此类框架。这种方法使用的图形结构类似于Ehresmann的草图解释的模糊集宇宙。这种方法利用草图的能力，集成数据的确定性和不确定性的结构。选择这一策略，我们也试图利用图形语言，在一般的范畴理论和用于草图定义，特别是如何适合于推理的问题，结构描述和任务规范和任务分解。
英文摘要：
The integration of knowledge extracted from different models described by domain experts or from models generated by machine learning algorithms is strongly conditioned by the lack of an appropriated framework to specify and integrate structures, learning processes, data transformations and data models or data rules. In this work we extended algebraic specification methods to be used in this type of framework. This methodology uses graphic structures similar to Ehresmann's sketches interpreted on a fuzzy set universe. This approach takes advantages of the sketches ability to integrate data deterministic and nondeterministic structures. Selecting this strategy we also try to take advantage on how the graphic languages, used in Category theory in general and used for sketch definition in particular, are suited to reasoning about problems, to structural description and to task specification and task decomposition.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
