博文
Application of Definition of Data, Information and Knowledge
|
The Application of Definition of Data, Information and Knowledge in Artificial Intelligence System
Yucong Duan
Benefactor: Shiming Gong
AGI-AIGC-GPT Evaluation DIKWP (Global) Laboratory
DIKWP-AC Artificial Consciousness Standardization Committee
World Conference on Artificial Consciousness
World Artificial Consciousness Association
(Email：duanyucong@hotmail.com)
Catalog
3 The distinction and application of conceptual space, semantic space and cognitive space
4 Professor Yucong Duan's emphasis on the distinction between concept and semantics
4.1 Definition and treatment of concepts
4.2 Definition and treatment of semantics
5 Comparison with other definitions and models
5.3 Polanyi's tacit knowledge theory
6 Detailed case of technology realization
6.1 Application in Natural Language Processing (NLP)
6.1.1 Realization of Concept Space
6.1.2 Realization of Semantic Space
6.1.3 Realization of cognitive space
6.2 Application in Knowledge Mapping
6.2.1 Realization of Concept Space
6.2.2 Realization of Semantic Space
6.2.3 Realization of Cognitive Space
6.3 Application of Intelligent Decision Support System
6.3.1 Realization of Concept Space
6.3.2 Realization of Semantic Space
6.3.3 Realization of Cognitive Space
Under the background of current research and application of artificial intelligence (AI), processing, interpretation and trust are always the core challenges. The DIKWP model proposed by Professor Yucong Duan provides a solid theoretical basis and methodological framework for constructing a treatable, interpretable and credible artificial intelligence system by distinguishing data, information and knowledge in detail and further distinguishing conceptual space, semantic space and cognitive space. This report will demonstrate the technical advantages of this model in detail, and show its application potential in artificial intelligence systems, especially the feasibility of technical realization.
2 Overview of DIKWP modelProfessor Yucong Duan's DIKWP model provides a high-order, dynamic and structured cognitive framework for cognitive subjects (such as human beings or AI systems) through in-depth definition and semantic differentiation of data, information and knowledge. The core of this model is to distinguish conceptual space, semantic space and cognitive space, so as to realize systematic processing, semantic association and cognitive understanding of data, information and knowledge.
Definition: Data is the original fact or observation record confirmed by the cognitive subject, which is classified and organized through the conceptual space to form a preliminary cognitive object.
Semantics: Data ensure its consistency in the cognitive process through semantic matching and association in semantic space.
Processing: Data form specific observation records and facts through thinking and classification in cognitive space.
Definition: Information semantically associates data with existing cognitive objects through the purpose of cognitive subject, identifies differences and forms new cognitive content.
Semantics: Information forms new semantic relations and information semantics through semantic matching and association in semantic space.
Processing: Information forms new semantic association and information through the purpose and cognitive process of cognitive subject in cognitive space.
Definition: Knowledge is a systematic understanding and interpretation of data and information through high-level cognitive activities and assumptions, forming a profound understanding and interpretation of the world.
Semantics: Knowledge forms systematic understanding and rules through semantic matching and association in semantic space.
Processing: Knowledge forms systematic knowledge and understanding through observation, hypothesis, abstraction, verification and correction in cognitive space.
3 The distinction and application of conceptual space, semantic space and cognitive space
Definition: Conceptual space is a space for cognitive subjects to communicate and recognize through natural language, symbols and other forms. Data, information and knowledge exist as concrete concepts in this space and are expressed through semantic networks and concept maps.
Technical realization: Through natural language processing (NLP), knowledge representation and semantic network construction, the system can understand and deal with complex concepts and relationships in the language. For example, use a graph database (such as Secondary) to build and manage concept maps.
Definition: Semantic space is the space where cognitive subjects understand and deal with the internal semantic relations of concepts. Data, information and knowledge understand and generate new knowledge through semantic matching, association and transformation in this space.
Technical realization: By constructing knowledge map, semantic search and reasoning engines (such as RDF and SPARQL), the system can generate new knowledge and understanding through semantic association and matching. For example, using knowledge map technology to realize semantic association and reasoning.
Definition: Cognitive space is the internal psychological space for cognitive subjects to think, learn and understand. In this space, data, information and knowledge form a profound understanding and explanation of the world through cognitive activities such as observation, hypothesis, abstraction and verification.
Technical realization: Through deep learning, reasoning and decision support systems (such as TensorFlow and PyTorch), the system can carry out complex cognitive activities and realize intelligent decision-making and action. For example, deep learning technology is used to simulate and realize cognitive activities.
4 Professor Yucong Duan's emphasis on the distinction between concept and semantics
Professor Yucong Duan particularly emphasizes the distinction between concepts and semantics, which is the core innovation of DIKWP model. By distinguishing conceptual space, semantic space and cognitive space, the model can process data, information and knowledge in multiple categories, thus achieving more efficient and accurate knowledge generation and understanding.
4.1 Definition and treatment of concepts
Definition: A concept is a concrete cognitive object or symbol expressed in the concept space, which is expressed through semantic network and concept map.
Technical realization: Using natural language processing technology and knowledge representation method, the system can understand and deal with complex concepts. For example, semantic networks and Ontology are used to define and manage concepts.
4.2 Definition and treatment of semantics
Definition: Semantics are internal relations and meanings generated by semantic matching and association in semantic space.
Technical realization: Using semantic search and inference engine, the system can generate new semantic connections and understandings. For example, semantic tagging and semantic search techniques are used to process and generate new semantics.
5 Comparison with other definitions and models
Definition: DIKW model defines data, information, knowledge and wisdom through hierarchy.
Contrast: DIKW model emphasizes hierarchy, but lacks deep distinction between semantics and concepts, while Professor Duan's definition realizes more efficient knowledge processing and generation through careful spatial distinction.
Definition: SECI model generates knowledge through the transformation of explicit knowledge and tacit knowledge.
Contrast: SECI model focuses on the dynamic transformation process of knowledge and emphasizes the interaction between explicit and tacit knowledge. Professor Duan's definition distinguishes conceptual space, semantic space and cognitive space, and realizes more detailed knowledge processing and generation.
5.3 Polanyi's tacit knowledge theory
Definition: Tacit knowledge is difficult to formalize and transmit, mainly through personal experience and skills.
Contrast: Polanyi's theory emphasizes the individuality and difficulty of transferring tacit knowledge, while Professor Duan's definition realizes the structured and formal expression of tacit knowledge through the fine distinction between semantics and concepts.
Definition: The Cynefin framework emphasizes the application and decision-making of knowledge in different situations.
Contrast: The framework of Cynefin emphasizes the situational adaptability of knowledge, while Professor Duan's definition realizes the systematic processing and generation of knowledge through the spatial differentiation of multiple categories, which is suitable for more complex cognitive tasks.
6 Detailed case of technology realization
6.1 Application in Natural Language Processing (NLP)
6.1.1 Realization of Concept Space
Technical method: Use semantic network and Ontology to construct conceptual model.
Implementation tools: graph database (such as Secondary) and ontology building tools (such as Protégé).
Application example: Building a semantic network in the medical field, the system can understand the relationship between "diabetes" and "insulin" and infer that "diabetic patients need to use insulin".
6.1.2 Realization of Semantic Space
Technical method: Use semantic search and reasoning engines (such as RDF and SPARQL) to realize semantic association and reasoning.
Implementation tools: Apache Jena, OpenLink Virtuoso.
Application example: enter "diabetes treatment method" in the semantic search engine, and the system can return relevant treatment methods based on the knowledge map, and relate them to specific drugs and treatment guidelines.
6.1.3 Realization of cognitive space
Technical methods: Deep learning techniques (such as TensorFlow and PyTorch) are used for language reasoning and decision-making.
Implementation tools: TensorFlow, PyTorch, Keras.
Application example: By training the deep learning model, the system can automatically generate personalized treatment plan according to the patient's symptoms and historical medical records.
6.2 Application in Knowledge Mapping
6.2.1 Realization of Concept Space
Technical method: Using graph database to construct and manage knowledge map.
Implementation Neo4j: Secondary, Amazon Neptune.
Application example: Build a knowledge map within the company, record the responsibilities of each department, the skills of employees and the progress of the project, and help the management to allocate resources and make decisions.
6.2.2 Realization of Semantic Space
Technical method: Using semantic search and inference engine to generate and manage knowledge.
Implementation tools: Apache Jena, GraphDB.
Application example: Through semantic search, we can find employees with similar project experience and form a new project team to improve the success rate of the project.
6.2.3 Realization of Cognitive Space
Technical methods: Deep learning and reinforcement learning techniques are used for knowledge reasoning and decision support.
Implementation tools: DeepMind, OpenAI Gym.
Application example: According to the project progress and feedback from team members, the system constantly adjusts the project plan and resource allocation to optimize the project implementation effect.
6.3 Application of Intelligent Decision Support System
6.3.1 Realization of Concept Space
Technical method: Use decision-making model and knowledge representation method to express and manage decision-making factors.
Implementation tools: decision support system (DSS) software, rule engine (such as Drools).
Application example: the enterprise supply chain management system is constructed, which can represent the relationship and decision-making factors of each link in the supply chain.
6.3.2 Realization of Semantic Space
Technical method: Semantic search and reasoning engine generate new decision-making knowledge.
Implementation tools: Oracle Semantic Technologies, AllegroGraph.
Application example: Through semantic search and reasoning, the system can find potential risks and opportunities in the supply chain and provide optimization suggestions.
6.3.3 Realization of Cognitive Space
Technical methods: High-order cognitive activities and deep learning techniques are used for complex decision-making reasoning.
Implementation tools: IBM Watson, Microsoft Azure AI.
Application example: the system makes real-time supply chain optimization decisions according to market changes and internal data of enterprises, and improves the competitiveness and response speed of enterprises.
Professor Yucong Duan's DIKWP model provides a solid theoretical basis and methodological framework for building a treatable, interpretable and credible artificial intelligence system by distinguishing data, information and knowledge in detail and further distinguishing conceptual space, semantic space and cognitive space. Compared with the traditional knowledge model, the DIKWP model shows unique advantages in dealing with complex systems and abstract concepts, especially in the following aspects:
Efficient data processing: Through the classification and organization of concept space, the system can efficiently process and manage a large amount of data.
Accurate semantic association: Through semantic matching and association in semantic space, the system can generate and understand new semantic associations and realize accurate knowledge generation and reasoning.
Deep cognitive understanding: Through high-order cognitive activities in cognitive space, the system can carry out complex learning, reasoning and decision-making, and realize a deep understanding and explanation of the world.
The application cases of DIKWP model in natural language processing, knowledge map and intelligent decision support system show its wide applicability and feasibility of technical realization. Through continuous research and improvement of this model, we can expect the birth of a more intelligent, efficient and credible artificial intelligence system, which will bring far-reaching influence and change to all walks of life.
Professor Yucong Duan's DIKWP model is not only of great significance in theory, but also shows strong technical advantages in practical application. Future research can further explore the application of this model in more fields and promote the development and progress of artificial intelligence technology.
https://blog.sciencenet.cn/blog-3429562-1435408.html
上一篇：Definition of Data, Information and Knowledge in DIKWP Model
下一篇：Semantic Mathematics: P = NP? from EXCR and ESCR Theories
