Ontology-Based Knowledge Management System - Research Paper Example

Proposal

Ontology- Based Knowledge Management System and Application

Project Background

The body of knowledge is growing at a very fast pace as a result of rapid advancement of information technology and its application. This growing multi-domain body of knowledge can be used in remarkable ways that can help the enterprises can competitive advantage. This explains a corresponding increase in the development of technologies directed at knowledge management (KM). The formal specification is conceptually modelled as ontology [1]. KM system based on ontology is acquainted with the tendency to adequately integrate related resources, quick search of knowledge, and deter the irrelevant knowledge. Until now, some large enterprises have used KM, including the United States NIST knowledge base project [2], the European WIST engineering knowledge management projects [3], and some domestic theories of the like suggested by Wujiang, that offers the KM system in a structured manner [4]. Another ontology-based KM system has been offered by Ye Ronghua [5]. All of them have attained some knowledge of effective realization, attainment, organization, and sharing, though there is room for more. Moreover, the common effect of the dynamic participation level not being sufficient is there, also the knowledge regarding the way the system can be accessed and maintained is closed, and all this comes in the way of making progress and breakthroughs.

Proposed Project

This paper divides the knowledge process into acquiring, storing, linking with knowledge mining, and knowledge representation. The root of this process is ontology. This paper develops a framework for KM system that is based on ontology, for which mining knowledge offers the information, that is structured and shared as ontology-based knowledge, with a view to reaching a shared ontology which can be comprehended by the computers as well as by the humans. An efficient interface of knowledge retrieval can help people have better understanding of different concepts. This encompasses innovation of knowledge as well as the use of management tools that ensure timely preservation and updating of the innovation knowledge. In this paper, a knowledge management system’s framework is created based on ontology through detailed analysis and retrieval of knowledge. The partly open system realizes the retrieval of knowledge in an efficient manner on the basis of improved organization and expression of knowledge. The next step is to study the way to gain the related knowledge through processing tools and auxiliary means, the way to detail the modeling of ontology, and the way to effectively process dynamic management using the tools of management.

Project Benefits

The openness of this management system allows it to constantly increase the body of tacit knowledge and efficiently polymerize the explicit knowledge. As a result, the designers will have increased support for innovation through a more efficient knowledge management and application system.

Procedure for the Proposed Work

The first step is the form of knowledge representation. The complexity and diversity of knowledge makes its structured expression quite complicated. Knowledge representations exist in a variety of forms presently, but there is dearth of a structured and mature knowledge representation method. Moreover, a unitary knowledge base establishment cannot address the requirements of all kinds of innovative designs since the knowledge in many fields is presently integrated and interrelated. This paper proposes a multi-ontology bases model regarding the organization knowledge. The purpose of this model is to create a unitary knowledge base, employ the fundamental ontological characters and relations explained before that define the interrelation of multi-ontology bases, and eventually create a network of knowledge as depicted by Fig. 1.

Fig.1. Ontology knowledge organization model

There is a tree of ontology relation in logical layer. Ontology relations interlink each concept. This layer comprises each concept and relation network of the knowledge base. Between each concept exist four elements-properties which are instances, axioms, relation and attribute sets. The upper structure of the library of ontology relations contains ontology layer. It makes a network of knowledge by supporting the design process including the equipment of process, techniques, and rules in process planning aided by computer, and linking of the function similarity by ontology relations while basing on the constructing model concept of multiple knowledge bases. The identity of knowledge networks as the main concept of related fields is widely accepted and the ontology knowledge network is the part of multi-ontology bases. Multi-ontology bases can be established with related domain knowledge in this paper in terms of developing ontology and logical layers. For example, it can contain different components such as the base of domain, principle, and integrated knowledge. All ontology knowledge bases are related by restraint bottom among the different concepts and describe the concept, object, and semantic relation of the areas on the basis of ontology. This helps in getting different design granularity of specifications, expression, integration, and sharing.

KM is directed at serving as a bridge between knowledge and the needing people so that they can reach optimal decisions through it. In addition, a major problem in KM is the retrieval of knowledge that serves as the hinge while connecting knowledge to people. It is imperative that the retrieval of knowledge is based on the organization of knowledge because organization patterns resolve retrieval patterns, and the process is opposite to that of knowledge organization [11]. Therefore, this paper makes use of domainontology-based pattern of organization, using ontology research and the representation and organization of knowledge so as to formulate the retrieval mode like depicted in Fig. 2.

Fig.2. Model of knowledge retrieval base on ontology

There are four parts of this retrieval mode, namely interlinkage of knowledge, ontology, resources of knowledge and matching retrieval. The primary function of interlinkage of knowledge is to get into the process of matching retrieval by means of retrieval passages, selecting the right entrance (auto retrieval and self-defined retrieval), that looks for the nodes and concepts of knowledge in the related fields, getting the right matching results and passing them on to the user. The foundation of retrieval of knowledge is the resources of knowledge based on ontology, which mainly creates the difference between the retrieval mechanism of knowledge and other common systems of information retrieval. Besides, it is system model’s core. Everything ranging from the analysis of retrieval and the handling of results to the process of retrieval of knowledge matching, marking of resources of knowledge, and index founding is based on the related ontology knowledge.

High-efficient retrieval of knowledge depends on a successful strategy and technique of retrieval. This paper formulates two paths of retrieval, namely auto retrieval and self-defined retrieval, so that the users at various levels can select the suitable methods of retrieval in accordance with their individual demands. Users can define knowledge aspects in the self-defined retrieval process using the concept navigation, the concept relations chart, and the classification navigation as the system uses ontology to search the related concepts with respect to the standard questions. Ontology can be conceptualized as categories, that can depict the fluctuation relations to the user in the interaction layer between computers and users. The view of listed ontology can be chosen by the users. Users are matched with the necessary knowledge as the users specify ontology in a choosing range. Such a search mode containing flexibility can address the needs of most users and offer them guidance in direction.

Automatic retrieval is another kind of retrieval path which is extended type retrieval, and expands the concepts. Automatic retrieval is done on the basis of the links between ontology concepts and semantics. Relations exhibited in the domain ontology’s hierarchical structure are used to increase the results of retrieval. Superclass concepts replace the user-demanding concepts or attribute values replace the specific attribute value, thus limiting this retrieval mode’s constraints. When a subclass concept replaces user-demanding concepts, deeper and more semantic expression forms and concepts can be obtained with the interdisciplinary concepts, including a variety of knowledge fields and concept-relative knowledge fields, so that the fields of interdisciplinary knowledge can be expanded.

Description of the Completed Project

The most important purpose of having a knowledge management system (KMS) is to improve the process of sharing of knowledge within an enterprise. Acquisition of knowledge is, hence, not just the first, but also the most basic requirement of KM [6]. To make the reuse of knowledge simple, KMS lays out items clearly, that exist in a knowledge base containing the information in unstructured, semi-structured, and structured forms. To realize this function, the KMS is separated into three components, including knowledge acquisition, knowledge storage, and knowledge reuse. The whole ontology based process is related through the mining, representation, and connection of knowledge. Fig. 3. Shows the framework of KMS design based on ontology.

Fig. 3. Ontology-based knowledge management system framework

Knowledge acquisition is the process of abstraction based on ontology’s concept. The process changes the knowledge stored in various forms into a structured format. Mining of knowledge helps realize the acquisition of knowledge so that the different sources of knowledge like documents, data bases, Web and applications can be stored in the knowledge base after the knowledge discovery system (KDS) has processed them. Other sources of knowledge, like all types of information forums, and user feedbacks including tacit knowledge are initially stored in the transit depot. These sources are stored in the knowledge base after the managers have effectively sorted them out. This makes knowledge acquisition the process of construction of knowledge and not the process of its conversion. Storage of knowledge is a process for transforming the information existing in unstructured and semi-structured forms into a structured format and placing it in the knowledge base. Knowledge sources give the metadata and the objects of knowledge are marked in virtue of the standards of metadata and ontology [7]. The base of ontology consists of relations of classified ideas of the domain objects of knowledge and various other concepts. The metadata required by KMS is stored in the metadata base which primarily facilitates efficient searching of the knowledge objects. Knowledge base and data base are a combination of relations of knowledge objects and the semantic metadata information. The divided stratum control on related knowledge regarding ontology and metadata base is the premise as well as the basis on which knowledge search and speculation is obtained efficiently.

Knowledge reuse, achieved through connections, is the process of use of knowledge in the systems of application. The knowledge search engine helps users find the relevant contents in the various strata. This is like getting knowledge by a pull method. Additionally, KMS yields the related knowledge based on immediate requirements and personal preference of the user. Knowledge managers regulate knowledge in the base, renew it, and save it in time, under the friendly management interface condition. This enables the systems to have dynamic participation ability rather than staying limited to static usage and close maintenance.

Ontology is a category of philosophy. The community of artificial intelligence defined ontology in a new way with the advancement in artificial intelligence. Presently, the definition of ontology given by Studer [8] that calls it a shared conceptual model of formal specification is widely accepted. Ontology is practically described in terms of an array of five components including class or concepts, relations, functions, axioms, and instances [9]. The fundamental component in the five-body array of ontology is relation. The fundamental element on which the structure of KM in ontology is based is the development of a good base of related domain ontology. Ontology relations reflect the limitations and a new link between concepts. The nature of these relations can be hyponymy, non modified, causality, appositive, synonymous or composition [10]. All types of nodes of knowledge can be coherently related through the different relations so that they comprise an ontology based knowledge relations network, and the correct node of knowledge can then be identified through the relative path. This paper divides the knowledge relations network into two components, namely main and secondary relation of ontology. The main ontology relation elaborates the various terms regarding particular fields and relations of ontology as they exist among all terms whereas the secondary ontology relation elaborates other fields’ external terms relating the terms of relations of main ontology. Use of main ontology relation together with secondary ontology relation can not only help develop a good knowledge organization, but also help in making the process of knowledge retrieval quick and efficient.

Feasibility of the Project

This system can be applied in the field of mechatronics, through more research on the establishment of knowledge management system’s framework based on ontology, the structuring and representation of knowledge, and retrieval. Different knowledge acquisition tools can be used to research the related knowledge fields to mine, sort, and create the related knowledge base. A successful system of KM cannot be constructed without a well-designed ontology. The hierarchy that exists between vocabularies widely accepted in this field are clearly defined in this paper, and the paper adopts a top-to-bottom approach. This implies that the top-level concepts are listed first, and the sub-categories are established gradually thereafter. Electromechanics, for instance, can be separated into different classes including engineering, power, driver, and performance, and the sub-categories can be spread gradually. A conceptual graph can be made spreading by analogy, that contains the domain knowledge with relativity; finally, a model of ontology can be constructed through defining the classes’ attributes. Application is the value of knowledge base of ontology. The link between questions and knowledge in base is fundamental to the retrieval of knowledge. As explained before in the model of retrieval of knowledge, users selecting the self-defined retrieval enter the interface.

Feasibility of the Project

Review of the existing literature is suggestive of the fact that there is inadequate relationship between technologies and knowledge processes to sufficiently elaborate their employment in the corporate dynamics. This imparts a need to associate the technologies with the initiatives of KM on the basis of well-developed knowledge strategies and KM approaches. There is room for more research to explore the exemplary KM initiatives that can help connect a variety of fundamental intents of knowledge such as customer profiling, new product development and improvement in operational processes, to the specific approaches and technologies of KM.

Schedule for the Proposed Project

Project Activity

Timeline

Writing Project Proposal

Dec 2017

Literature Review

1 Jan 2018 – 31 March 2018

Reaching Specific Project Goals

1 April 2018 – 30 April 2018

Creating questionnaire

1 May 2018 – 31 May 2018

Obtaining Test Data

1 June 2018 – 31 July 2018

Data Analysis

1 August 2018 – 31 August 2018

Writing Final Report and Preparing Presentation

1 September 2018 – 30 November 2018