Fuzzy Decision Support System for Science and Technology Project Management ^†

Abstract

To improve the accuracy and scientific of science and technology project management, a fuzzy decision support system was developed in this study. We designed the overall deployment architecture of the system, which consists of the system access layer, system core layer, system service layer, and basic platform layer. A Web server was used to reduce the response time of the system. The indices of science and technology projects were sorted by using the fuzzy decision support process and the expert’s weight matrix. To improve evaluation accuracy, a program and the storage process of the results were established at each stage of the evaluation. The developed system spent less time querying evaluation results. The query error rate was low, indicating improved efficiency of science and technology project management.

1. Introduction

Scientific and technological innovation is essential for national development. As a main carrier of scientific and technological innovation, scientific and technology projects directly impact the progress of national scientific and technological endeavors [1]. Since reformation and opening up, China has progressed in science and technology project management. The 20th National Party’s Congress proposed the reform of the science and technology system to innovate science and technology by optimizing the management of science and technology projects. Science and technology projects involve various research and development activities including funds, personnel, and procurement. Their management faces challenges of risks, complexity, and innovation, which makes it difficult to promote and manage science and technology projects [2]. In particular, reviewing the results of experts requires the following in science and technology project management: multi-objective hierarchical decision-making [3], the determination methods of index weights or quantitative scores [4], and consistent checks of individuals and groups [5].

A decision support system (DSS), developed in the late 1970s, is a computer-based human–machine interactive system that assists decision-makers in solving semi-structured and unstructured problems by using data and models [6]. As the most active and rapidly developed decision analysis method, the fuzzy decision-making method satisfies resource constraints and diversity of project criteria [7], helping organizations to solve challenges of multi-objective decision-making of science and technology projects [8]. Therefore, we applied the fuzzy decision support system to science and technology project management to improve the management ability for science and technology projects.

2. System Hardware

2.1. System Architecture

The developed system has an architecture composed of the framework and key components to ensure efficient and reliable operation. The framework needs to be determined according to system scale, number of scientific and technology projects, reliability, and security. We designed the deployment architecture of the science and technology project management system as shown in

Table 1. Deployment architecture of developed system.

System Access Layer	Web Management Interface		Data Interface
System core layer	Applications management	Initialization management	Project management
	Achievement registration	Technology contract	Scientific report
System service layer	System management	Platform management	Database
Basic platform layer	Software environment
	Hardware platform

.

The system architecture consists of the system access layer, system core layer, system service layer, and basic platform layer. The basic platform layer is composed of a hardware platform and software for system operation. The system service layer is composed of system management, platform management, and a database. The system management layer provides personal information and account management. The platform management layer conducts information and operation maintenance, and the database is used to store data and logs. The data are backed up and used for recovery and mining [9]. The system core layer links the process management of science and technology projects, including application, project approval, project management, achievement registration, technical contracts, and scientific reports. The layer is also responsible for centralized storage, analysis, and processes. The project management layer conducts allocation, contract, change, implementation, and interim acceptance [10], which is the core process. The system access layer is composed of a Web management interface and a data interface to facilitate data integration with various platforms [11].

2.2. Layer Design

To improve the effectiveness of project management, the data of science and technology projects must be evaluated in the system user layer. Feedback and results from project leaders and managers are transmitted to the database for storage. Traditional science and technology project management systems need to install project evaluation applications in the system core layer, which increases the workload of the system and leads to installation problems due to the incompatibility of the system for data collection and processes. Therefore, we set up a Web management interface between the system core layer and the system user layer. The system user layer uses programs and communication protocols for direct evaluation. The Web server transmits the evaluation results to the system core layer. In this process, the system user layer does not need to install additional operating software, which reduces the workload of the system core layer effectively as well as the response time. If the user logs in to the server, corresponding operations automatically start, which reduces the cost of scientific and technology project query and evaluation. There is no limit to the location of users, which improves the work efficiency of scientific and technology project management.

3. System Software

3.1. Program Design

Traditional science and technology project management systems construct the judgment matrix of projects. However, the structure is vague, which leads to a high error rate. To increase the accuracy of project evaluation, the fuzzy decision support process was applied in this study to science and technology project management using the expert weight matrix. The process of the evaluation of the projects is as follows.

• Step 1: Establish an evaluating index system for the full life cycle of science and technology projects, which includes guideline evaluation, project approval review, progress assessment, and performance evaluation (

  Table 2. Full life cycle evaluating index system of science and technology project.

  First-Level Index	Secondary-Level Index
  Guideline evaluation (Q1)	Policy consistency (Q11)
  	Procedure fairness (Q12)
  	Content rationality (Q13)
  Project approval review (Q2)	Innovativeness (Q21)
  	Feasibility (Q22)
  	Risk (Q23)
  	Research basis (Q24)
  Progress assessment (Q3)	Schedule (Q31)
  	Quality standard achievement (Q32)
  	Well organized (Q33)
  Performance evaluation (Q4)	Achievement level (Q41)
  	Comprehensive benefits (Q42)

  ) [12]. The score ranges from 1 to 100. A beneficial situation is essential for the implementation of the results. The expected effect is presented in the output with an emphasis on economic, ecological, and social benefits [13].
• Step 2: Construct the weight rule of experts for science and technology projects. An evaluation committee is formed with three or five experts, and weights are determined. The leader of the evaluation committee is selected from the database of science and technology projects and is informed one day in advance. The leader of the authority of one vote veto on evaluating the results of science and technology projects [14].
• Step 3: According to the weight of the second-level index of science and technology projects, fuzzy decision support is used to sort factors in evaluation criteria to establish a fuzzy evaluation model (Equation (1)) [15].

  $$\mathrm{P}={\displaystyle \frac{\mathsf{\alpha }\times \mathrm{G}}{\sum _{i=0}^n{\mathsf{\omega }}_{\mathrm{m}}{\mathrm{Q}}_{\mathrm{i}\mathrm{j}}}}$$

  (1)

  where α represents the ambiguity of a single index, and G stands for the index data set.
• Step 4: The initial iteration process of evaluation is obtained by combining the fuzzy evaluation model and index data set, and results are normalized using Equation (2).

  $$\mathrm{Z}=\mathrm{P}\times \mathrm{G}$$

  (2)
• Step 5: Repeat Step 3 until the complete evaluation of all science and technology projects.
• Step 6: The final evaluation score is calculated in the life cycle stage.

The evaluation process of science and technology projects is shown in Figure 1.

3.2. Storage of Evaluation Results

After evaluating the projects, the results are stored in the database using the designed storage program in this study. User IP addresses are varied during the evaluation of projects. Therefore, the IP address is obtained first, and the evaluation result is displayed in a text box to establish a connection to the IP address of the user terminal. By doing this, the query rate is increased. The data format of the results is set to any byte. After receiving the results, the system service layer stores them in a unified array. Each time the results are received, the system service layer saves them as an element of an array. The storage process of the evaluation results is shown in Figure 2.

At the same time, the document name is stored in the interface of the project management system when storing evaluation results. For effective data storage, the function of changing the storage path is also used. To change the storage path, a new path can be chosen to assign the new data input path in a text box. Finally, the evaluation results are input in each life cycle stage. In addition, the developed system sets a query time, according to project number, project name, application unit, project host, award, and others for simple fuzzy search. It assists managers in managing two projects only in a year. The function of research query for projects is also provided in this system [16].

4. System Testing

4.1. Testing Environment

To verify the effectiveness of the project management system based on fuzzy decision, tests were conducted using Microsoft Programming, as shown in

Table 3. Testing environment of science and technology project management system.

Item	Type	Applicable Environment
Explorer	IE 11 above	User Side
Database	SQL Server 2024	Database Management
Web Server	Windows Server 2022	Web Server
Operating System	Windows 2022	System Deployment
Processor	Intel Core i9 8950hk	Information Processing, Program Running Execution Unit
Simulation Platform	Matlab 2024	—
Programming Language	Vc++ 6.0	Program Development

.

4.2. Method

It is necessary to refine the scale of science and technology project evaluation to improve reliability, reduce errors in project evaluation, and increase the evaluation accuracy of science and technology projects [17].

• Step 1: Develop an evaluation plan. To test the performance of the fuzzy decision support-based project management system, the advantages of this system are reflected. The query response time and query accuracy rate of evaluation results are selected as performance verification indicators.
• Step 2: Technology preparation. Three systems with different data amounts are loaded in the Matlab 2024 simulation platform, and relevant index information is collected on the background of the operating system intelligently.
• Step 3: Analysis of testing results.

4.3. Results and Analysis

Firstly, the query response efficiency of the evaluation results was calculated. The response efficiency is determined by the response time of the query process. The shorter the response time of the query process, the higher the efficiency of this system. Statistical results are shown in Figure 3. The greater the total amount of data and the number of evaluated scientific and technology projects, the longer it takes for the system to conduct evaluation queries. When the amount of data is ${10}^3$ bits, the query time is between 1.5 and 10 s; when the amount is $2\times {10}^3$ bits, the time is between 2.1 and 15.5 s; when the amount is $3\times {10}^3$ bit, the time is between 3.5 and 22 s. The developed system completed the query of evaluation results quickly, and the query process was completed within 22 s, indicating that this system has a high efficiency of evaluation query response.

We tested the query accuracy rate using random sampling times, and the testing results are shown in

Table 4. Testing results of query accuracy rate of evaluation results.

Number of Scientific and Technology Project	200	400	600	800	1000
Correct query times of evaluation results	49	48	66	64	80
Error query times of evaluation results	1	1	2	2	3
Accuracy (%)	96	98	97	97	96

. The query accuracy rate of the evaluation results reached 98% at the highest and 96% at the lowest. The system showed a higher query accuracy rate of evaluation results, indicating that this system has a strong effect.

5. Conclusions

We designed a science and technology project management based on the fuzzy decision support system. Based on the system architecture and the science and technology project management server, the evaluation procedure of the project was designed based on fuzzy decision support, and the storage process of evaluation results was established. The response time of the evaluation results query of this system was short, and the query accuracy of evaluation results was high, which indicates that the science and technology project management system designed in this study is highly effective.