Chinese name of basic introduction: Ontology mbth: Ontology first use of ontology: German scholastic scholar Guo Clanyue Time:17th century philosophy concept, research process, other fields, building ontology, introduction, steps, basic problems, overview, ontology history, concept, object, connection, entity and environment, existence and philosophical concept have different definitions of ontology. Generally speaking, the ontology used in philosophy before Marxism can be divided into broad sense and narrow sense, and philosophy after Marxism can be integrated into practice. Broadly speaking, it refers to the ultimate nature of all reality, which needs to be understood through epistemology. Therefore, the ultimate essence of studying all reality is ontology, and how to know it is epistemology, which is called ontology and epistemology. In a narrow sense, there are studies on the origin and structure of the universe, and in a broad sense, there are ontological studies on the nature of the universe. The former is cosmology and the latter is ontology, which is called ontology and cosmology. Marxist philosophy does not take the methods of ontology and epistemology, or ontology and cosmology, but uses dialectical materialism to explain the whole problem of philosophy. (Excerpted from Dictionary of Foreign Philosophy edited by Feng Qi) The study of "ontology" in the research process has its origin in the history of Greek philosophy. From the school of Miletus, early Greek philosophers devoted themselves to exploring the "primitive" (Greek arche, translated as "primordia"), the most basic element of all things. The study of this "primitive" became the predecessor of ontology, and gradually approached the discussion of existence. Later, parmenides profoundly put forward that "there is nothing outside, there is something, so there will be no non-existence". And think that existence is eternal, only thinking is the same with it, and only thinking can achieve this truth; From the point of view, there is existence and non-existence, and existence is not one, so it is extinct. Parmenides's exposition of existence establishes the basic direction of ontological research: the "existence" divided by "existence" can only be explored from thinking to transcendental field, but not from feeling to experience; Because of its absolute universality and originality, there can only be one "yes" found in the transcendental field. However, only Socrates and Plato can really understand this point, and Greek philosophers of the same period have more or less ignored it. So, for example, although atomists also distinguish between truth knowledge and dark knowledge and recognize the difference between thinking and feeling, it is extremely vague whether the "source" they seek can be obtained from experience, so they actually do not distinguish between transcendence and experience. In Socrates' dialogue without final conclusion, the possibility of obtaining truth through empirical induction has been eliminated; In Plato's idealism, the "idea" of transcendental world is obviously the basis of truth. In ancient Greek and Roman philosophy, the study of ontology is mainly to explore the origin or matrix of the world. Philosophers of various schools try to attribute the existence of the world to some material and spiritual entity or some abstract principle. Parmenides put forward the only unchangeable primitive "existence", which made the study of existence the theme of this period. Aristotle believes that the main object of philosophical research is entity, and the problem of entity or ontology is about essence, phase and individual things. He thinks that the philosophy of studying entity or noumenon is the first philosophy above all other sciences. Since then, the study of ontology has turned to explore the relationship between essence and phenomenon, * * and supernormal, and general and individual. In modern western philosophy, Descartes first called the first philosophy of studying entity or ontology "metaphysical ontology". During the 17~ 18 century, Leibniz and his successor Wolff tried to establish a complete set of metaphysics about general existence and the nature of the world, that is, an independent ontological system. Wolff regards generality and universality as the essence and reason of detachment from individual, singleness and independence. On the one hand, Kant believes that it is impossible to establish an abstract ontology, which can only study the universal essence of things and the difference between material existence and spiritual existence; On the other hand, ontology is replaced by a transcendental philosophical system separated from epistemology. Hegel put forward the unified principle of ontology, epistemology and logic on the basis of idealism, and constructed the logical system of dialectical development of existence itself from the concept of pure existence. In modern western philosophy, some schools (positivism, analytical philosophy, philosophy of science, etc. ) Against any metaphysics and ontology, some people try to re-establish ontology about existentialism, such as Husserl's transcendental ontology, Heidegger's basic ontology and Hartmann's critical ontology. They often use extrasensory intuition to establish conceptual systems, and their views are idealistic or agnostic. Ontology research In China's ancient philosophy, ontology is called "root theory", which refers to the theory of exploring the root causes and basis of the emergence, existence, development and change of all things in the world. Ancient philosophers in China generally attributed the origin of the universe to something completely different from the universe, which can be roughly divided into three categories: (1) substances with no fixed form, such as "Qi"; (2) abstract concepts or principles such as "nothing" and "reason"; (3) Subjective spirit, such as "mind". These three viewpoints belong to simple materialism, objective idealism and subjective idealism respectively. In the study of the history of China's philosophy, some scholars use the word "ontology" to refer to the idealistic theory of finding the basis of the existence of the material world outside the material world, such as Wang Bi's "Gui Hu" in the Wei and Jin Dynasties. Ontology in other fields is a philosophical concept and a philosophical problem to study the essence of existence. In recent decades, this word has been used in the computer field, playing an increasingly important role in artificial intelligence, computer language and database theory. However, so far, there is no unified definition and fixed application field of ontology. The definition given by Gruber of Stanford University has been recognized by many colleagues, that is, ontology is an accurate description of conceptualization (Gruber, 1995), and ontology is used to describe the essence of things. In implementation, ontology is a detailed description of conceptualization, and an ontology is often a formal vocabulary, whose core function is to define the professional vocabulary of a certain field or field and their relationships. This series of basic concepts, like the cornerstone of a building, provide a unified understanding for all parties to communicate. Under the support of these concepts, the efficiency of searching, accumulating and enjoying knowledge will be greatly improved, and it is also possible to realize the real reuse and enjoyment of knowledge. In this sense, XML is a typical application of ontology theory. Ann. An xml document is a tagged vocabulary. Ontology can be divided into four types: domain, generality, application and representation. Domain ontology contains related knowledge of a specific type of domain (such as electronics, machinery, medicine, teaching), or related knowledge of a certain subject or course; Universal ontology covers several fields, which is usually called core ontology. Application ontology contains all the knowledge needed for modeling in a specific domain; Representation ontology is not limited to specific fields, but also provides entities that describe things, such as "frame ontology", in which the concepts of frame and slot are defined. It can be seen that the establishment of ontology has a certain level. In the field of teaching, if the concepts, terms and their relations in a certain course are regarded as a concrete application ontology, then the concepts and characteristics of * * * in all courses have certain universality. Ontology, a philosophical category, has been given a new definition by the field of artificial intelligence, which has been introduced into information science. However, the understanding of ontology in information science is gradually developed and matured. 199 1 year, Neches et al. first gave the definition of ontology in information science: "Give the basic terms and relations that constitute the vocabulary in related fields, and define the extension rules of these words by using these terms and relations." Later, in the fields of information system and knowledge system, with more and more people studying ontology, different definitions appeared. 1993 gruber defined ontology as "a clear specification of conceptual model". In 1997, bost was further improved as "formal specification of * * * enjoyment concept model". Studer and others have conducted in-depth research on the above two definitions, and think that ontology is a clear formal specification for the conceptual model of * * *, which is also the unified view on the concept of ontology at present. The definition of ontology by Studer and others contains four meanings: conceptual model, explicit, formal and * * * sharing. "Conceptual model" refers to the model obtained by abstracting the related concepts of a phenomenon in the objective world, and its meaning is independent of the specific environmental state; "Clear" means that the concepts used and the constraints for using these concepts are clearly defined; "Formalization" means that ontology is readable by computer, that is, it can be processed by computer; "* * * Enjoyment" means that ontology embodies the knowledge recognized by * * * and reflects the concept set recognized by related fields, which is aimed at groups rather than individuals. The goal of ontology is to capture the knowledge in related fields, provide a common understanding of the knowledge in this field, determine the vocabulary in this field, and give a clear definition of the relationship between these words (terms) and vocabulary from different levels of formal models. Although there are many different ways to define ontology, from the perspective of connotation, different researchers have a unified understanding of ontology, and they all regard ontology as a field (the scope of a field can be a specific application or a broader scope). Semantic basis of communication (dialogue, interoperability, enjoyment, etc.). ) between different internal subjects (people, machines, software systems, etc.). ), that is, ontology provides a kind of knowledge. Moreover, this knowledge provided by ontology is mainly for machines. Machines can't understand the semantics expressed by natural languages like humans, and current computers can only treat texts as strings. Therefore, when discussing the ontology in the computer field, we should discuss how to express * * * knowledge, that is, the formalization of concepts. Ontology in the semantic web? The emergence of ontology research hotspots is also directly related to the proposal and development of semantic web. Semantic Web is another concept put forward by Tim Berners Lee. Tim Berners Lee believes that the current network is for people to read and understand. As an increasingly huge archive medium, it is not conducive to the automatic processing of data and information. The new generation of semantic Web should not only bring semantic content to human beings, but also bring semantic content to computers (information agents), so that computers (or information agents) can "understand" the web content and realize the automation of information processing. He believes that the semantic web is not another web isolated from the current web, but an extension of the current web. In the semantic web, the semantics of information is well defined, which can better promote the cooperation between computers and people. In order to realize the function of semantic web, it is necessary to provide a structured semantic description mechanism that computers can understand and a series of reasoning rules to realize automatic reasoning. The challenge of Semantic Web is to provide a language that can express data and inference rules in data. It needs this language to apply the rules currently existing in knowledge representation system to the Web. In Tim Berners Lee's semantic web framework, there are several key components. They are XML, RDF and Ontology. XML allows users to define their own file types and arbitrary complex information structures, but XML is only grammatical, and it cannot explain the semantics of the defined structures. The importance of XML in the semantic web is closely related to the fact that XML is a carrier language and the naming mechanism of XML. In Tim Berners Lee's view, semantic description is carried out through RDF. Two characteristics of RDF make a special contribution to this: (1)RDF is a ternary structure composed of resources, attributes and attribute values. This ternary structure is similar to the relationship between subject, predicate and object in a sentence. An RDF statement describing resources is as effective as the sentence "What kind of attributes does something have?". It can show an assertion of the existing state of things and express the knowledge that computers need to deal with in most cases. (2) Another important feature of RDF is that the resources, attributes and attribute values that make up RDF must be identified by URIs (Uniform Resource Identifiers). Because RDF uses URIs to encode information, it means that any resource, attribute and attribute value referenced by RDF are predefined fuzzy concepts. RDF can express declarative sentences, and the subject, predicate and object are all determined by URIs, so it has the characteristics of semantic expression. But the requirements of the semantic web are far more than that. The semantic web also needs to add logical functions: the semantic web needs to be able to use rules to reason, choose action routes and answer related questions. Ontology is the basis of logical reasoning in semantic web. ? Semantic Web researchers also believe that ontology is a standardized document, which formally defines the relationship between words. For the semantic web, the most typical ontology has a classification system and a series of reasoning principles. Among them, the classification system defines the categories of objects and the relationship between categories. The class/subclass relationship between entities is of great value to Web applications. In ontology, you can also add attributes to a class to define more category relationships. These category relations provide the basis of reasoning. With the help of reasoning rules in ontology, semantic Web application system can provide stronger reasoning ability. For example, a rule can be added to a geographical ontology, "If a city code is related to a provincial code and an address uses a city code, then this address is related to the corresponding provincial code". Through this rule, the program can infer that the Document Information Center of Chinese Academy of Sciences is located in Zhongguancun, which should be located in Beijing. In order to describe ontology, researchers of ontology semantic web developed RDFS on the basis of RDF. RDFS can describe the relationship between concepts with the help of several predefined words (such as RDFS: resource, RDFS: class, RDFS: attribute, subclass of RDFS:, RDFS: subpropertyof, RDFS: domain, RDFS: scope). In order to describe ontology more conveniently and comprehensively, W3C is actively promoting the application of OWL(Web Ontology Language) on the basis of RDFS and with the help of related research of DAML and OIL. Since February 10, 2004, OWL has become the standard recommended by W3C. In order to find relevant information in a certain research field, biologists often spend a lot of time. To make matters worse, different biological databases may use different terms, just like some dialects, which makes information search more troublesome, especially irregular machine search. Gene ontology (GO) is a project initiated to solve this problem. The basic concept in gene ontology is $ term. Each entry in GO has a unique digital label, such as GO: nnnnnnn, and a $ term name, such as Cell, Fibrogrowth Factor Receptor Binding or Signal Transduction. Every $ term belongs to an ontology, and there are always three ontologies, namely, molecular function, cellular components and biological processes. A gene product may appear in more than one cell component, or it may play a role in many biological processes and play different molecular functions in them. For example, the gene product "cytochrome c" is described as "oxidative educational activity" by molecular functional terminology. Biological process terms are used to describe "oxidative phosphorylation" and "inducing cell death". Finally, its cell components are "mitochondrial matrix" and "mitochondrial inner membrane". $ term in gene ontology has two interrelationships, namely, relationship and partial relationship. Is_a relation is a simple inclusion relation. For example, a is _a B, which means that a is a subset of b, for example, the nuclear chromosome is _a chromosome. The relationship of part_of is a little more complicated. C part_of_D means that if C appears, it must be a part of D, but C does not always appear. For example, the nuclear part_of cell, the nuclear must be a part of the cell, but some cells have no nuclear. The structure of gene ontology is a directed acyclic graph, which is similar to a classification tree, but the difference is that a $ term can have multiple parent nodes in the structure of gene ontology. For example, the biological process $ TERM "hexose biosynthesis" has two parents, which are hexose metabolism and monosaccharide biosynthesis, because biosynthesis is a metabolism and hexose is a monosaccharide. Gene ontology uses the definition in Oxford Dictionary of Molecular Biology (1997), and also refers to Switzerland -PROT, PIR, NCBI·CGAP, EC… The established standards are not the only standards (which GOC has always emphasized), and naturally it is not stipulated that every researcher must follow this control word set system. The dynamic structure adopted is to connect the ontologies in series with the neork of DAGs (directed acyclic graph) to form a hierarchical tree, that is, the relationship between "being a" and "being a part of it" mentioned above. Because GO is an integrated classification system, its three main ontologies are independent, but whether it is the original design of GOC or our use, there is still a certain process relationship. In the annotation process of gene/protein or ontology, firstly, the components and elements involved in cells are considered, secondly, the molecular function of the components/elements is exercised at the molecular level, and finally, the biological process in which the molecular function directly participates can be presented. Because this is a labeling process with feedback mechanism, and the whole system is dynamically open and updated in real time, it has the ability of error correction to some extent. TAMBIS project is the only system that realizes the integration of information resources at the conceptual and contact levels. But there are other related plans under study. For example, BioKleisli (Department of Computer Science, University of Pennsylvania) realized the integration of multiple data sources by using Mediator technology, and the subsequent K2/Kleisli system also realized OLAP (Online Analytical Processing) by using data warehouse. DiscoveryLink(IBM Research Institute) realizes the integration of information sources based on wrapper/mediator, and puts forward query decomposition and cost-based optimization strategy. TAMBIS (Department of Computer Science, University of Manchester) realizes the integration of information sources based on wrapper/mediator, uses CPL language in BioKleisli as the query language, and gives the method of query optimization. TaO's ontology provides domain knowledge for users to browse and query. TINet(GSK Company and IBM Research Institute) defined the object view of data source based on multi-database middleware OPM(Object-Protocol Model), and its CORBA (common object request broker architecture) server made the wrapper of data source easier to expand. But they all have certain defects. The query ability of BioKleisli system is weak, and no query optimization strategy is given. The number of data sources integrated by TAMBIS system and DiscoveryLink system is relatively small, and the latter does not use domain knowledge in query processing, so query decomposition is not considered from the semantic point of view. The query processing ability in TINet system is not strong. At present, the outstanding problem is the lack of the ability to select the best retrieval members and dynamically generate the best retrieval scheme in database queries, especially when there is overlapping information in multiple databases. Moreover, the existing work mainly focuses on data integration, and rarely considers service integration. The system supporting the comprehensive integration of data and services is not perfect. Therefore, this will also be an important direction for the future development and promotion of Go. Introduction to Ontology Construction In practical application, ontologists, knowledge management, artificial intelligence, information science (library science) and even any department or field with a large amount of information that needs to be classified can become the application objects of ontology. The basic element of ontology is $ term)/ concept, and vocabulary constitutes homogeneous classes and subclasses, and then a simple ontology is formed by adding appropriate relationships between each class and concept. Concepts and classes are used to express words themselves, while relationships provide mappings for words and add constraints to make them conform to the actual situation. Protege, the most commonly used software for ontology construction, was developed by Stanford University. It was first used in the medical field, and then gradually extended to other fields. The software itself is also open source software. Steps The simple steps to build ontology are: 1. List the terms involved in the research topic. 2. Summarize and modify terms according to their inherent attributes and exclusive characteristics, and establish a class and hierarchical classification model for terms. 3. Add a relationship to connect terms and taxonomy 4. Finally, according to the need, in Protege, you can use its attached functions and plug-ins to export ontology in the form of words and graphics, and the format can be freely selected. The fundamental question of ontology is: "What is the original classification of existence?" Philosophers of different schools have different interpretations of this problem. This problem makes it necessary to use ontological knowledge in theology, library science and artificial intelligence research, and it is embodied in cosmology, morality and aesthetics. In addition, the questions raised by ontology include: What is existence? What is a real thing? What are the basic properties of an object? What's the difference in determining an object? What is the basic definition of "physical existence"? What are the characteristics of objects? What are the characteristics of existence? When does an object cease to exist? Not just change. Why can objects exist relative to nothingness? Historical ontology can be traced back to Plato and Aristotle in ancient Greece. Aristotle once defined ontology as "the science of studying the existence of objects". Specifically, it studies the classification of objects, that is, under what circumstances, an object can be defined as "existence". Comparative theory includes "universal" problems and "specific" problems, concepts, objects and connections. Many questions can be asked about existence: "What is existence?" "What is it?" "What am I?" For ontology, the most basic thing is to understand what is an object, what is an idea, and the relationship between them. Descartes' idea of "I think, therefore I am" in the Enlightenment initiated the basic problem of ontology. Descartes didn't study it deeply, and he didn't think it was necessary to study it deeply. He thought that "I think therefore I am" proved the existence of God, and later theologians also thought: "Who can apply this sentence to everything in general?" Of course, there is only "God". In the 20th century, when the new discoveries of philosophy of mathematics, philosophy of science and elementary particle physics broke the so-called boundaries of many substances before, people were no longer satisfied with theological views. In different periods of the 20th century, there were different schools of metaphysical subjectivism, objectivism and relativism. Postmodern entity philosophers try to redefine the above problems through philosophical actions in different environments, and mainly rely on the latest research results of biology, ecology and cognitive science to understand the cognitive status of animals in natural and artificially provided environments. The change of proposition relative to different environments makes existence more difficult to define. If people say, "A is B", "A must be B" or "A was B", what does it mean? Some philosophers advocate removing the word "yes" in English and replacing it with an abstract meaning to avoid confusion; Other philosophers try to understand the deep meaning and usage of words; Martin heidegger wants to distinguish the meaning of "being" and "object". Existentialists believe that "existence" is the most basic concept, and many things can be said to be "yes", while the verb "yes" has many uses, so it is easy to be confused, so there are many kinds of existence. Simply put, ontology is to explore whether there is an abstract foundation behind everything in this world that does not depend on the real world. Both spirit and matter have their own abstract foundation. Simply put, metaphysics refers to the real world that we can feel, and metaphysics refers to the reasons behind the perceptual world, which are abstract and intangible and exist as the foundation of the perceptual world. Ontology is the metaphysical basis for exploring the metaphysical world.