Knowledge fusion method based on fuzzy set theory

Abstract

The discussion of knowledge management and technological innovation has never stopped, and the discussion of the relationship between the two has not only important practical significance but also profound theoretical significance. The purpose of this paper is to study a new method of knowledge fusion from the perspective of science and technology philosophy. From a new perspective, this paper analyzes defects of subjective tendencies, decision-dependent partial attributes of knowledge source existed in practical applications of management field and proposed a knowledge fusion method based on fuzzy set theory. This paper firstly explains the characteristics of knowledge sources, and transforms knowledge into a new knowledge layer through combining multi-source knowledge, then improves the connotation, level and self-confidence of knowledge, finally improves the ability of the system to accomplish tasks and goals. Then combine the fuzzy set theory with the knowledge fusion algorithm reasonably and effectively, and obtain the results of knowledge fusion by using evidence synthesis and decision rules, so as to make up for the lack and defects in the knowledge fusion process and solve the uncertainty problem in knowledge reasoning. Finally, through the practical example, merged the fuzzy set theory proposed in this paper into knowledge fusion to deal it, obtain a kind of processing of fuzzy set theory, forming a knowledge fusion method based on fuzzy set theory. Based on fuzzy set theory, obtain the observation results of knowledge fusion algorithm combined with the various warning models, then to discuss and analyze the enterprise warning problem deeply. Therefore, the examples and simulation results show that the advantages in practicality and versatility of knowledge fusion method proposed through fuzzy set theory is higher than the common knowledge fusion method. The method used in the production of manufacturing products can help manufacturing companies improve development quality of product and shorten development cycle of product. Moreover, in the product design industry, it has verified that knowledge fusion can promote the dissemination of knowledge in the field of knowledge management, which helps to share and reuse design knowledge, reduce difficulty of development and improve efficiency of development.

Keywords

Knowledge management information fusion fuzzy set theory fuzzy petri net knowledge update

1 Introduction

With the leaping advancement and development of science and technology and the arrival of the knowledge economy, the foundation of knowledge production and development slowly replaces the original traditional materials such as capital, natural resources, and labor. This material is not a key factor for the rapid development of enterprises [1]. Therefore, in today’s economic globalization, how can we significantly enhance the technological innovation capability of many enterprises, and gradually become a stage that must be experienced in enterprise development, and also the key to development. The knowledge economy is a new form of economy. The so-called knowledge economy has many proposals different from the previous ones. In traditional industries, it is a pillar industry, and scarce natural resources are mainly dependent on the new economy. This is the first pillar industry and high-tech industry. The main intellectual resources are therefore the sustainable development of the economy. Facing the ever-changing environment, how to effectively, quickly acquire, develop, integrate, and innovate knowledge and cultivate the learning ability of enterprises is the key to business success. Therefore, it can be understood that if enterprises want to acquire these necessary factors based on the market, they must rationally manage the knowledge resources and promote the integration and development of multiple knowledge [2]. In this case, knowledge management came into being. As a new fusion method for knowledge fusion, current knowledge management can not only broaden the new application scenarios of knowledge management, but also promote the innovation and development of technology. Therefore, we can clearly notice the necessity and importance of the development of knowledge fusion, and play a very important role in technological innovation. Then, what is the specific relationship between knowledge management and technological innovation ability, and what is the value thinking between them? It will be the problem that this article tries to solve [3].

The current economic quality is one of the conditions for measuring social development and human existence. It occupies a very important position in human daily life and affects the healthy and sustainable development of society. Economic development not only affects the transformation and development of ideas, but also promotes the advancement of science and technology, production and education [4]. In the era of knowledge-based economic development, economic development is not only dependent on traditional resources, but also on the development of knowledge and the integration of knowledge [5]. The current economic development is mainly promoted by both technical knowledge and management knowledge. The former has a great role in the advancement of science and technology, and the latter can significantly promote the improvement of management level. It is precisely because of the mutual development of the two. Knowledge has gradually become one of the important conditions for promoting the rapid development of the economy, and it is also the foundation and development motivation of the development of the knowledge economy. While the continuous development of knowledge enriches the self-innovation, multi-source knowledge is constantly developing and renewing self-generation. The continuous integration of multi-source knowledge and new knowledge can not only increase the intrinsic, hierarchical and credibility of knowledge itself. And it enables decision makers involved in knowledge fusion to use the new knowledge of fusion to make faster, more effective and correct judgments [6]. The current fusion algorithm as one of the implementation price adjustment of multi-source knowledge is also one of the key technologies of multi-source fusion, which will directly affect the effect and development of multi-source knowledge fusion. Compared with information fusion technology, knowledge fusion is not mature from theory to practice. At the same time, knowledge fusion is more application-oriented, and integrated decision-making and decision-making are receiving more and more attention [7]. According to the important process in the process of knowledge fusion, this paper seeks the commonality and individuality of its fusion, draws on the idea and development process and method of multi-source information fusion, and carries out the secondary development in the field of knowledge fusion using the excellent method and method of information fusion [8]. When directly observing the ever-changing market environment and understanding the speed of new products in the market environment, how to control product design from the perspective of knowledge management and knowledge fusion, to achieve timely and accurate design knowledge and experience, accurate reuse, Has become a major problem in the manufacturing industry. To further improve design efficiency and quality, and shorten the product development cycle is one of the conditions that need to be considered. At present, product design is more and more valued by enterprises. Through advanced design technology, enterprises can improve product quality, improve product performance, and reduce product cost [9] to increase product circulation speed. At the same time, product design has also changed from experience design to knowledge design. The application knowledge of product design is no longer a single subject knowledge, but an interdisciplinary comprehensive knowledge [10]. Therefore, it is of great significance to study multi-source knowledge fusion. Early scholars mostly used it as part of knowledge engineering and combined it with other related content [11].

The terminal studied in this paper is mainly to find multi-source knowledge with expandability and complex connotation in the field of knowledge management, and integrate and utilize the value uncertainty of multi-source knowledge. Then, the paper analyzes the different ways of knowledge sources, discusses the ways and development conditions of knowledge fusion level, and the status of knowledge processing. It proposes a method to distinguish multi-source knowledge. Based on the idea of information fusion, a multi-source knowledge fusion algorithm based on fuzzy set theory is proposed. The multi-source knowledge fusion model based on Petri net is established, and the processing steps of the model are given. The multi-source knowledge fusion algorithm based on fuzzy set theory based on Petri net criterion is discussed. Then the corresponding algorithm is implemented for the given algorithm implementation. Process and analysis of the performance of common knowledge fusion algorithms. Finally, the effectiveness of the knowledge fusion algorithm application simulation experiment is verified for the enterprise fault early warning problem.

The first chapter is an overview. The main introduction is a brief introduction to the definition and advantages of knowledge fusion methods based on fuzzy set theory. The background, research significance and work of this paper are presented. The second chapter introduces the method, introduces the basic methods of information fusion, introduces the detailed principles of several typical information fusion methods and introduces the basic fuzzy set theory, and summarizes the fuzzy set theory multi-source information fusion. The third chapter is an experiment that details the source of the experimental data and the type of data, and informs the experimental environment in which the experiment is performed and the steps that the experiment needs to perform. The fourth chapter is the experimental analysis, the experimental examples are unfolded, and the experimental content, experimental process and experimental details are introduced. The fifth chapter analyzes and summarizes. This chapter mainly analyzes and summarizes the experiments. On the basis of careful summarization and summary of the research, this paper points out the gains and losses of this study.

2 Proposed method

2.1 Related work

(1) Status of knowledge fusion research

The source of the concept of fusion of knowledge is accompanied by the development of knowledge engineering. At first, many researchers were more inclined to knowledge fusion as part of knowledge engineering, and the two were jointly analyzed and discussed [12]. With the development of knowledge engineering, knowledge fusion has gradually become one of the research directions. The research on knowledge fusion as the research theme has mushroomed. At the same time, the adoption of related computer technologies and the research process of participation have also increased [13]. At the same time that knowledge fusion has blossomed at home and abroad, domestic research has received high attention and received more and more attention. There is no recognized definition of knowledge fusion in foreign countries, and the research focus is not the same, but the description of knowledge fusion has basically the same meaning [14]. In the past five years, China’s research on knowledge integration has gradually begun. Although it started late, it caused great concern. Through the survey of the website of the National Natural Science Foundation of China and the website of the National Philosophy and Social Science Planning Office, the project support of the China National Knowledge Integration Research Fund was obtained [15]. Through the retrieval of knowledge networks, intellectual property, Zhou fang and other databases, it is found that the authors of knowledge fusion research are also members of the above projects, and most of the papers published are from the above projects. For example, Zhou Fang The information fusion method is introduced into the field of knowledge fusion based on the idea of information fusion, and three knowledge fusion algorithms based on Bayesian criterion, d-s evidence theory and fuzzy set theory are proposed [16]. A knowledge fusion algorithm based on fuzzy set theory is proposed. The idea is to fuse the fuzzy data located on [0, 1] according to the given fusion function, and then process the newly generated data [17].

(2) Research status of knowledge management

The concept and practice of knowledge management is a new thing, and it is not the same as knowledge fusion for knowledge management. However, it is precisely because of people’s research purposes and perspectives that the understanding and interpretation of knowledge management are different. Well-known knowledge management expert Burch H believes that knowledge management is a kind of catering measures that organizations face in the face of changing environment [18]. The trend of knowledge economy and knowledge management is the first to propose the concept of knowledge management. The father of management master Carl Latilla VM has been researched since the 1970 s. In 1986, another management master Drucker believed that enterprise knowledge activity is “knowledge management” [19]. The famous Japanese management scholars Calabrese G and Mastroberardino P put forward the “knowledge transformation theory” in the book “Research on the Mechanism of Knowledge Innovation in Organizations”, which is the mutual transformation theory of explicit knowledge and tacit knowledge. At present, in the field of knowledge management, from the published literature, it mainly discusses the knowledge expression of management, as well as the principles and methods of knowledge management, organization and systems, strategies and goals, focusing on the knowledge management system framework and architecture research, have not seen Discussion knowledge fusion and integration of multiple sources [20].

2.2 Knowledge fusion algorithm based on fuzzy set theory

Knowledge fusion algorithm is the key technology evolved in the process of knowledge fusion. The quality of the algorithm directly affects the quality of knowledge fusion. The requirements of knowledge fusion algorithm selection are mainly attributed to the task requirements encountered in the process of knowledge fusion, the level of knowledge fusion and so on. The fuzzy set theory has the advantages that other algorithms can’t match for some problems that are not easy to determine, the information is easy to conflict, and the realization of subjective and objective integration. The fuzzy set theory was first proposed by Zadeh, which further relaxed the constraints of probability axioms and has a broader application space.

(1) Fuzzy set theory

Fuzzy logic is to some extent an extension of Boolean and Boolean logic [21]. The difference is that it extends Boolean binary logic to continuous value logic by introducing a member function called muon. The membership function is a continuous function with a value interval of [0, 1] used to quantitatively describe the ambiguity or inaccuracy of the information. The formula is as follows: $μ_{A} : \begin{matrix} U \to [0, 1] \\ u \mapsto μ_{A} (u) \end{matrix}$ (1)

Where U is the domain, and A is the fuzzy set in U represented by a real number. For u ∈ U, function value μ_A (u)

Become the membership of element u on set A.

A classical set is a special case of a fuzzy set, that is, a feature set of a classical set is a special membership function. Based on the computational properties of classical set feature functions, Zadeh introduces the inclusion relations, equality relations and operations of union, intersection and complement of fuzzy sets as follows.

Let U be the domain, A and B are the two fuzzy sets on U.

If ∀u ∈ U, A (u) ≤ B (u), then A is called a subset of B (A is included in B), which is: $A \subseteq B \Leftrightarrow A (u) \leq B (u), \forall u \in U$ (2)

If SAD, i A and B are equal, and it is recorded as A = B, that is: $A = B \Leftrightarrow A (u) = B (u), \forall u \in U$ (3)

The sum of A and B, and its membership function is: $(A \cup B) (u) = A (u) \lor B (u)$ (4)

The intersection of A and B, and its subordinate function is: $(A \cap B) (u) = A (u) \land B (u)$ (5)

The complement of A is recorded as A′, and its membership function is: $A^{'} (u) = 1 - A (u)$ (6)

For the two fuzzy sets, the union, intersection, and complement operations can generalize the union, intersection, and complement operations. That is, set T to any given set of indicators. ∀t ∈ T is the fuzzy set on U, which is {A_t} _t∈T, and is recorded as $\underset{t \in T}{\cup} A_{t}$ membership function is: $(\cup A_{t}) (u) = \underset{t \in T}{\lor} A_{t} (u)$ (7) $(\cap A_{t}) (u) = \underset{t \in T}{\land} A_{t} (u)$ (8)

(2) Fuzzy logic knowledge fusion

Knowledge fusion based on fuzzy logic mainly includes three steps: fuzzy knowledge fusion, fuzzy fusion and clarity:

Fuzzy knowledge. First, a mathematical model of the specific problem is established, and a limited reference set is given based on prior knowledge. Then, based on the designer’s experience, a membership function is constructed on the reference set of each knowledge base. Finally, the knowledge provided by the knowledge source is blurred by the membership function. In order to simplify the calculation, a trigonometric function or a trapezoidal function is often used as a membership function.

Fuzzy fusion. According to a certain fusion function, the fuzzy value between multiple [0, 1] is fused to obtain a new value between [0, 1], which is the result of fuzzy fusion. The fuzzy fusion process of N knowledge sources is expressed as: $μ (θ) = f (μ_{1} (θ), . . ., μ_{N} (θ)), θ \in υ$ (9)

The fusion function f is a fuzzy logic described mathematically, or it can be a fuzzy logic rule described in a table or language.

A jealousy. Sharpness refers to the process of restoring knowledge data from the fuzziness obtained by fuzzy fusion. For example, for feature fusion, the ambiguity in the interval [0, 1] is mapped to the feature space, and then the knowledge feature data is restored.

Fuzzy integral is a nonlinear function defined on a set of fuzzy measurements. With h : X → [0, 1], the fuzzy integral on A ∈ X is defined as: $\int A^{h (x) ° g (•) = \sup_{0 \leq α \leq 1} [\min (α, g (A \cap a_{α}))]}$ (10)

If X represents a set of knowledge sources and h is a membership function of the knowledge source, then the fuzzy integration of Equation (2) completes the knowledge fusion process.

2.3 Fusion algorithm based on fuzzy evidence theory

The important role of the knowledge fusion method is to integrate multi-source knowledge information through reasoning, and provide a new method and decision-making tool for product design. Therefore, knowledge reasoning is a very important step in the process of knowledge fusion. In knowledge reasoning, although the ontology itself has the function of reasoning, the complexity of the ontology reduces the efficiency of the fusion algorithm. At the same time, in product design, especially in conceptual design, often encounter uncertain or incomplete knowledge, so it is impossible to use ontology to achieve effective knowledge fusion. Therefore, other knowledge fusion algorithms are needed to achieve local fusion of uncertainty knowledge.

The main methods of uncertainty reasoning are subjective Bayesian reasoning, evidence theory, fuzzy logic reasoning, neural network and genetic algorithm. Comparing these kinds of inference methods, the calculation formulas of subjective Bayesian method are mostly derived on the basis of probability theory and have a solid theoretical basis. It requires domain experts to give the prior probability P (H) of H while giving the rules, which is difficult to do in practice. Although neural networks work well when the knowledge information set contains a large amount of noise input information, it does not guarantee optimal convergence results due to the lack of ability to interpret its own behavior, especially the ability to interpret the content being learned. Evidence theory only needs to be satisfied, and the probability axiom system can handle the “I don’t know” caused by uncertainty. The conclusion of knowledge may be a more general hypothesis, which is easy to be semantically from different expression domain experts. Knowledge, without being constrained by a single element, represents the most obvious level. Therefore, this paper chooses evidence theory as the main method of uncertainty knowledge reasoning.

Knowledge integration plays an important role in knowledge service and is one of the important contents of library information work. By reading a large amount of literature and summarizing, comparing and analyzing, it is found that there are some differences in the research on knowledge integration at home and abroad, and domestic research still has the following deficiencies.

(1) The fusion algorithm of knowledge fusion is not precise enough. Although China is relatively mature in the research of key technologies of knowledge fusion, the fusion algorithm is diversified and comprehensive, but its accuracy and efficiency need to be further improved.

(2) The structure of knowledge fusion system is not mature, and a special, unified and perfect knowledge fusion system model has not yet been formed. In the process of architectural research, although domestic scholars tried to get rid of the limitations of foreign research and establish a system model that conforms to the field of library information in China, a general model has not yet been established.

(3) Knowledge integration has not yet been fully applied to the knowledge services of libraries and information. At present, domestic scholars apply knowledge fusion to the field of computer network structure, military and other knowledge management. In the future, it will be used in manufacturing, healthcare, environment, economics, technology and other fields. The library information field should become the leader in the application of knowledge fusion in the field of knowledge services.

In view of the insufficiency of domestic research in the field of knowledge fusion, future research in China should focus on the following four aspects: establishing a common knowledge fusion architecture and promoting relevant research in various disciplines; applying knowledge fusion to more interdisciplinary subjects, fully The key role of knowledge fusion in the realization of knowledge services; in the big data environment, with the increase of modern knowledge and knowledge base, knowledge fusion should clearly define the goal and give full play to its role. In addition, domestic research on knowledge fusion started late and researchers are limited. Therefore, in this field, domestic scholars should cultivate more scientific and technological innovation talents to study knowledge integration in a deeper, more specific and more systematic way.

3 Experiments

3.1 Data collection

The application of knowledge fusion method based on fuzzy set theory in enterprise fault identification is simulated, and the effectiveness of knowledge fusion method is verified [22]. In this example, five listed companies A, B, C, D and E were selected. And these five companies are basically engaged in the product manufacturing industry, the following is a detailed introduction of each company: Company A is mainly engaged in aerospace products and components, as well as aviation technology development, consulting and services, but no ST records. Company B is mainly the development of formula computer software and hardware and computer hardware and software sales. The company does not have ST records. Company C is similar to Company B. However, Company C mainly conducts secondary development and sales of terminal products such as computer parts. The company does not have ST records. Company D is mainly engaged in the development of information technology and data products, and has strong business capabilities. Sales services are spread all over the country. The company does not have ST records. E company is mainly for the development of medical equipment and the development and sales of medical product technology, but E company is the only company that has a company’s settlement loss this year. The company has ST records.

According to the annual report publicly released by the five companies throughout the year, they are extracted from relevant financial data.

3.2 Experimental environment

In the experimental environment of this paper, the experiment uses an improved petri net, which requires the system to have higher operating efficiency and friendly interaction interface. Tools that make it easy to implement complex data structures and easily connect to databases. Therefore, choose mat lab 2016a as the development tool of the system, and choose python as the development language.

3.3 Experimental steps

The knowledge fusion method based on fuzzy set theory mainly includes three experimental steps:

(1) Uncertainty modeling: The identification framework that mainly acquires the basic probability distribution function or the trust function, through data mining and testing, etc., based on the provided data knowledge source and prior knowledge, to provide evidence for subsequent mathematical description of uncertainty.

(2) Fuzzy set theory combination. The combination of fuzzy set theory is the process of knowledge fusion. The combination of fuzzy set theory may be based on the basic combination rule of the basic probability assignment given in equation (10), or other combination rules may be derived from equation (10).

(3) Optimal decision making. The optimal hypothesis is determined by the base probability distribution obtained by the fusion. One of the most commonly used criteria is to choose the assumption with the strongest support as the optimal judgment. The maximum support criterion of the judgment process includes two processing steps: support degree calculation and optimal judgment. The support degree calculation is based on the basic probability distribution function of trust and the rationality. The upper and lower bounds constitute an uncertainty interval; the optimal hypothesis consists of Determination of the maximum lower bound and minimum range criteria for uncertain intervals.

4 Discussion

4.1 Performance analysis of three kinds of knowledge fusion algorithms

(1) As shown in Table 1, according to the Bayesian method, the reliability and confidence of the three fusion methods of D-S theory and fuzzy set theory knowledge show that the three methods are popular decision fusion processing methods. In the collection or institutional framework for research objects, based on mathematical theory, they all consider the uncertainty of knowledge acquisition. By means of fusion processing, we can easily distinguish the uncertainty of knowledge, so that we can judge the quality of the fusion result from the reliability and confidence of the fusion result.Comparison of three knowledge fusion algorithms as show in Table 1 and Figure 1.

Table 1
Comparison of three knowledge fusion algorithms

Bayes D-S Fuzzy set

Reliability 78.9 85.6 89.8

Confidence 84.6 83.6 88.9

	Bayes	D-S	Fuzzy set
Reliability	78.9	85.6	89.8
Confidence	84.6	83.6	88.9

Fig.1

Comparison of three knowledge fusion algorithms.

(2) However, each algorithm has the advantage that other methods are irreplaceable, but at the same time there are inherent disadvantages that are difficult to overcome. Therefore, this experimental experiment mainly selects three algorithms from the three aspects of real-time, accuracy and robustness from the knowledge fusion algorithm, from the existing specific conditions and the existing data, as shown in Figure 2.

Fig.2

Comparison of the advantages of the three algorithms.

4.2 Enterprise application experimental results and comparison

(1) Layer fusion methods are often used to discuss the discrimination of knowledge fusion in enterprise application failures, that is, to observe the same enterprise in different ways, from different perspectives, the knowledge that will be acquired to obtain richer and more accurate Knowledge and higher confidence and make more accurate judgments. Here we choose the z-score model and the Logit model as the source of knowledge. The discriminative indicators of these two models are considered in terms of capital liquidity, business capability, capital structure and growth capacity, but the discriminating indicators used in the survey are not the same, which means that the observation angles are different. The above data is calculated using the z-score model of two knowledge sources, as shown in Table 2 and Figure 3.

Table 2
Two knowledge source Z-score model calculation table

Company X1 X2 X3 Z_score Y1 Y2 Y3 Logit

A 0.38 0.1 0.16 2.45 0.32 0.2 0.09 0.0013

B 0.34 0.11 0.78 2.56 0.5 0.09 –0.05 0.0023

C 0.16 0.26 0.27 1.34 0.72 0.1 –0.2 0.0034

D 0.43 0.52 0.65 1.78 –1 –0.18 –0.07 0.0025

E 0.26 0.01 –0.14 0.69 0.66 0.27 –0.05 0.0022

Company	X1	X2	X3	Z_score	Y1	Y2	Y3	Logit
A	0.38	0.1	0.16	2.45	0.32	0.2	0.09	0.0013
B	0.34	0.11	0.78	2.56	0.5	0.09	–0.05	0.0023
C	0.16	0.26	0.27	1.34	0.72	0.1	–0.2	0.0034
D	0.43	0.52	0.65	1.78	–1	–0.18	–0.07	0.0025
E	0.26	0.01	–0.14	0.69	0.66	0.27	–0.05	0.0022

Fig.3

Knowledge source Z-score model calculation diagram.

(2) According to the annual report published by the five companies, relevant financial data is extracted from it, and the corresponding values are calculated by Logit model. The calculation results are shown in Figure 4.

Fig.4

Knowledge source Logit model calculation.

5 Conclusions

From the analysis and results of the experiment, it can be seen that the knowledge fusion method based on fuzzy set theory can obtain the maximum consistency of the two models Z-score and Logit in order to judge the success of the enterprise, so as to obtain accurate judgment results. As a result, the actual situation of the five companies in the amateur forecast is not much different, which proves that the research method proposed in this paper can play a role in the fault identification of the enterprise and is effective. However, this method requires a certain prior knowledge, and the fuzzy membership function has a great influence on the fusion effect. Moreover, from the above simulation experiments and results, it is not that the process of multi-source knowledge fusion from the same knowledge can support and complement each other, so that it can prove that the single-source knowledge discrimination is more reliable without multi-source knowledge source discrimination. Achieve better decision making. Fuzzy set theory can effectively process and process fuzzy uncertain information, complete fuzzy reasoning or obtain decision results. Applying it to practical cases, we can obtain the fuzzy reasoning ability of multi-decision systems, and further improve the decision-making credibility of information fusion systems. In this paper, the fuzzy set theory is used to solve some information fusion problems in multi-information fusion. This paper introduces the powerful comprehensive analysis ability of fuzzy comprehensive evaluation, and proposes an information fusion method based on fuzzy comprehensive evaluation to make up for the shortcomings of other information fusion methods with high computational complexity.

Through this research, it is found that the mathematical characteristics of fuzzy set theory and data thinking mode have a good effect on multi-source information fusion, and can be applied to the development process of information fusion, thus achieving good results of knowledge fusion and knowledge management. However, how to apply the fuzzy set theory effectively in the development field of multi-source information fusion is an urgent problem to be considered. How to overcome the shortcomings of other knowledge fusion methods in multi-source knowledge, such as large amount of calculation and unsatisfactory fusion effect, is one of the problems that need to be solved urgently in the future. The research on information fusion algorithms is endless. If you try the multi-fusion method, you can get a relatively efficient and rapid knowledge fusion method, which is also one of the theoretical directions for further research in the future. The field of knowledge fusion is an important branch of knowledge management and one of the important research fields. We are currently studying a new method of knowledge fusion from the perspective of science and technology philosophy, and analyzing the practical application of knowledge management from a new perspective. The existence of knowledge sources has the defects of subjective tendency and decision-dependent partial attributes, and proposes a knowledge fusion method based on fuzzy set theory to improve the scientific level of knowledge management. It is applied to the field of knowledge fusion-based management. It discusses the fusion, fusion modeling and decision fusion algorithms and fusion in hierarchical knowledge fusion processing. The main conclusions are as follows: To build a knowledge fusion model, it is necessary to consider the characteristics of knowledge sources and the task of fusion. Target, the complexity and real-time requirements of the fusion method, try to choose a simple fusion model to represent the complex knowledge fusion system. The Petri network is a powerful tool for asynchronous and concurrent systems for modeling and performance analysis. It can be used for static structural analysis and dynamic behavior analysis.

Knowledge fusion is a comprehensive and integrated multi-source knowledge, which is an effective tool and way of thinking. Through integration and processing, the connotation, quality and trust of knowledge can be improved and new knowledge can be generated. Compared with information fusion technology, knowledge fusion is more oriented to the application layer and decision makers, and at the same time, the processing method is still immature. Using information fusion, the knowledge fusion processing method and general model are discussed. Three multi-source fusion algorithms based on Bayesian rules, D-S evidence theory and fuzzy set theory are discussed. The corresponding processing steps are given. Construct a multi-source knowledge fusion processing framework. The work of this paper has a good reference value for further research on multi-source knowledge fusion processing technology.

Footnotes

Acknowledgments

This work was supported by the Soft Science Research Project in Shaanxi Province (No. 2019KRM047).

References

Baptista

, Vasconcelos

J.B.

and Silva

, The impact of perioperative data science in hospital knowledge management, Journal of Medical Systems 43(2) (2019), 41.

Moscoso-Zea

, Castro

and Paredes-Gualtor

, A hybrid infrastructure of enterprise architecture and business intelligence & analytics for knowledge management in education, IEEE Access 7(99) (2019), 1–1.

Saravanan

, Mohanraj

and Senthilkumar

, A fuzzy entropy technique for dimensionality reduction in recommender systems using deep learning, Soft Computing – A Fusion of Foundations, Methodologies and Applications 23(8) (2019), 2575–2583.

, Baldi

and Liu

, The non-smoothness problem in disturbance observer design: A set-invariance-based adaptive fuzzy control method, IEEE Transactions on Fuzzy Systems (99) (2019), 1–1.

Centobelli

, Cerchione

and Esposito

, How to deal with knowledge management misalignment: A taxonomy based on a 3D fuzzy methodology, Journal of Knowledge Management 22(3) (2018), 538–566.

Burch

, Moore

C.E.

and Burditt

J.A.

, Measuring the nutrition knowledge of weight management and diabetes risk in a low socioeconomic population, Topics in Clinical Nutrition 34(1) (2019), 47–56.

Lee

and Mcbride

, Extended object tracking via positive and negative information fusion, IEEE Transactions on Signal Processing 67(7) (2019), 1812–1823.

, Markov

and Matsui

, Articulatory and spectrum information fusion based on deep recurrent neural networks, IEEE/ACM Transactions on Audio, Speech, and Language Processing 27(4) (2019), 742–752.

Ichimura

and Tachibana

, Emotion orientated recommendation system for hiroshima tourist by fuzzy petri net, Communications in Computer & Information Science 244 (2018), 660–667.

10.

Chang

, Wu

and Chen

, Comprehensive risk assessment of deep water drilling riser using fuzzy Petri net model, Process Safety & Environmental Protection 117 (2018), 483–497.

11.

Letia

T.S.

and Kilyen

A.O.

, Using unified enhanced time petri net models for cyber-physical system development, IFAC-Papers Online 51(2) (2018), 248–253.

12.

Butts

, Higgins

and Dunbar

, The third time’s a charm: Psychometric testing and update of the atlanta heart failure knowledge test, Journal of Cardiovascular Nursing 33(1) (2017), 13.

13.

Tosi

, Lio

and Teatini

, Land subsidence in coastal environments: Knowledge advance in the venice coastland by TerraSAR-X PSI, Remote Sensing 10(8) (2018), 1191.

14.

Penninx

, Problems of and solutions for the study of immigrant integration, 7(1) (2019), 13.

15.

Prieto-Pastor

, Martín-Pérez

and Martín-Cruz

, Social capital, knowledge integration and learning in project-based organizations: A CEO-based study, Journal of Knowledge Management 22(8) (2018), 1803–1825.

16.

Rafatnia

, Nourmohammadi

and Keighobadi

, Fuzzy-adaptive constrained data fusion algorithm for indirect centralized integrated SINS/GNSS navigation system, GPS Solutions 23(3) (2019), 62.

17.

Wang

, Shan

and Zhao

, A novel algorithm for identifying key function nodes in software network based on evidence theory, International Journal of Software Engineering and Knowledge Engineering 29(03) (2019), 415–432.

18.

Burch

, Moore

C.E.

and Burditt

J.A.

, Measuring the nutrition knowledge of weight management and diabetes risk in a low socioeconomic population, Topics in Clinical Nutrition 34(1) (2019), 47–56.

19.

Latilla

V.M.

, Frattini

and Petruzzelli

A.M.

, Knowledge management, knowledge transfer and organizational performance in the arts and crafts industry: A literature review, Journal of Knowledge Management 22(4) (2018), JKM0820170367.

20.

Calabrese

and Mastroberardino

, How do knowledge and innovation work within organizations: The situationist view of the firm (SVF), International Journal of Innovation & Technology Management 14(05) (2017), 20.

21.

Sadatiseyedmahalleh

M.R.S.

and Measuring

, the safety by using the fuzzy logic methods inuniversitisains Malaysia, Revista Científica 25(4) (2015), 1–9.

22.

Xiao

and Ding

, Divergence measure of pythagorean fuzzy sets and its application in medical diagnosis, Applied Soft Computing 79 (2019), 254–267.