Architecture of a Knowledge-Based Education System for Logistics

Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 269)


Nowadays the technologies develop so fast that people cannot keep pace with the increasing logistics knowledge. In order to provide a professional platform of learning modern logistics, moreover, to shorten learning time and improve the learning efficiency, this paper introduces a knowledge-based education system which provides different user groups with well-formed learning contents. The architecture of the system as well as its four main modules is the focus of the paper. Additionally, the learning strategies and the mechanism of formulating learning plan are interpreted. It is a good foundation for the future knowledge-based education system in the logistics and supply chain management.


Education system Logistics SCM Knowledge base Knowledge-based education system 

337.1 Introduction

Nowadays, the economics develops increasingly fast and the demand on goods exchange grows also quickly. In order to adapt to different application environment in practice, logistics systems with more advanced technologies are deployed, algorithms with higher complication are used for the planning of logistics systems and numerous logistic strategies emerge. Under this circumstance, the data and information as well as the knowledge and experiences in the area of logistics increases continuously day by day.

Whilst data and information are surely without effort to be stored, the knowledge and experience differing from people are hardly to keep. A DIKW Chain is mentioned in [1] to reveal the differences between data, information, knowledge and experiences. Thus, it is necessary to find a proper solution to maximize the utilization of knowledge and experiences. In such sense, the knowledge-based system (KBS), which can collect, store and manage organized knowledge, is appropriate to support the purpose. In [2], several relevant technologies for modeling, managing and representing knowledge in KBS are investigated and presented.

The application of KBS has many successful examples in the area of medicine, agriculture and manufacturing already. In this paper, however, it focuses more on applying KBS in education. Knowledge and experiences should not merely be kept and used. More important, they should be passed on to the next generation. In [3], architecture of an E-learning system in manufacturing industry is introduced. With manufacturing training materials, scenario case base and problem solving methods being provided, the system educate the employees to ensure their continuous competence in handling their task. In addition, an educational system based on semantic web technology is written in [4]. This technology provides more adaptability, robustness and richer learning environments in building KBS. Moreover, a Web-based education system with the purpose of creating Web learning environment is proposed in [5]. The integrated tools with Web Services have a characteristic of powerful adaptability for management, authoring, delivery and monitoring of learning content. Furthermore, KBS is also applied in logistics. A knowledge-based logistics strategy system referred in [6] is designed for supporting the logistics strategy development stage by retrieving and analyzing useful knowledge and solutions.

However, there exists little research on applying KBS as educational system for logistics. Combined with the idea of customizing learning contents and process, as described in [7, 8], we strive to put forward a kind of KBS which serves to educate people with personalized logistics knowledge. Firstly, we focus on depicting the architecture of the knowledge-based education system for logistics in Sect. 337.2. Then in Sect. 337.3, the application process of the system will be elaborated with an illustration. Afterwards, characteristics of this system will be portrayed. At last, results of applying KBS with a case will be presented, in order to verify the feasibility of the system.

337.2 Knowledge-Based Education System for Logistics

The purpose of building Knowledge-based Education System for Logistics (KESL) is to educate different user groups with logistic knowledge. Based on a Knowledge Base (KB), which contains full of educational materials of logistics, we strive to organize these materials strategically and then present the knowledge in a way that learners can assimilate them on their best performance.

337.2.1 Architecture of KESL

The architecture of KESL consists of four modules and the system is divided into three layers, as illustrated in Fig. 337.1. Firstly, the internal layer is responsible for supporting the learning activities of interactive layer by obtaining, maintaining and organizing data and formulating instructional learning strategy. Additionally, the interactive layer mainly relates to user’s actions during learning. Moreover, the external layer is an indispensable ingredient, in which stand the system designer and supporting information from the Internet.
Fig. 337.1

The architecture of KESL

337.2.1.1 Module 1 Entrance

By Entrance, users can login and start their learning journey. In the first time, users will be asked to input their profiles. The configuration of their inputted information is the only identification for every individual. Consequently, the information will be saved in the Account Administration and used for formulate personalized learning plan. The required information from users is listed in Table 337.1.
Table 337.1

Options and attributes for classifying different groups of people



User groups

Student, teacher, company personnel, manager, specialist, common people, …


<18, 18–24, 24–30, 30–50, > 50

Be engaged in

Economics, finance, journalism, engineering mechanics, material science, supply chain management, …

Previous experience

No foundation, common knowledge, broad vision, specialized in (), …

Purpose of learning

Interests, brief understanding, basic skills, additional knowledge for main major, practical knowledge for work, teaching, know-how, …

337.2.1.2 Module 2 User Management

In this module, the Account Administration acts like a transit depot between users and message processing. On the one hand, it receives profile configurations from module 1 and obtains users’ feedbacks from module 4; on the other hand, it delivers feedbacks to Strategy Picker every time users add some feedback during learning. This operation repeats continuously and it forms a closed-loop feedback system, so that the Strategy Picker could be immediately updated and always ready to formulate an adapted learning plan. Moreover, to achieve effective knowledge storage, a multi-dimensional database (MDB) technology could be used, which is proposed in [9]. It facilitates the knowledge representation, navigation and maintenance with a high performance level. Furthermore, the process of storing knowledge is also a process of establishing clusters. It enables data mining for potential usage.

The History Navigator in this module relates to record the learning process for every user. These recordings are stored in personal account. They will be utilized afterwards in the process of reviewing at the beginning of the next round of learning.

Additionally, the Template Database stores kinds of interface styles. Users can define their own style and it will be saved in personal account.

337.2.1.3 Module 3 Knowledge Center

The Knowledge Center is composed of Strategy Picker and Knowledge base. It receives different information from other modules as constraints and according to them then formulates the appropriate learning strategy and pick up the corresponding educational materials.

Knowledge Base

The Knowledge Base (KB) consists of a database and its structure, by which different domains and layers of logistic knowledge are revealed, as illustrated in Fig. 337.2. In database, knowledge is collected from various ways: Encyclopedia for logistics, logistic Handbook, technical drawing and documents from logistic projects, scientific articles, curriculums provided by university, literatures, PowerPoint from internet and so on. Furthermore, all these collected knowledge are divided into five layers, as is shown in Table 337.2 and Fig. 337.2.
Fig.  337.2

Structure of the knowledge base

Table 337.2

Layers of the knowledge base and contents in each layer


Knowledge examples

General introduction

Terms about logistics and SCM, PPT and PDF files.

Logistics systems

Explanation of logistics systems with picture, video, 3D-animation., planning procedure, Best practices

Subsystems and equipment

Container and pallets, rack system, forklift, conveyors, sorter etc.

Functions and principles

FIFO, LIFO, location management, route control, transport regulations and roles, layout plan principles

Algorithms and methods

Performance calculation, dimensioning, simulation etc.

Additionally, in order to make sure that all the knowledge objects can be picked out appropriately, they must be assigned several attributes. These attributes are previously given by the knowledge base designer and will be matched with constraints from users. They can describe the location of every educational material, reveal a fuzzy relationship between knowledge objects or indicate the abstraction level of a knowledge object and so on.

Strategy Picker

Strategy Picker is actually an inference engine. It is responsible for formulating the most appropriate instructional learning plan. As already mentioned in former text, module 3 receives different information from other modules which contains constraint conditions. After the Strategy Picker has received those constraints, it starts to search for a basic learning strategy at first and then to collect needed knowledge through the basic learning strategy. According to different users, there are primarily three basic learning strategies applicable:
  • Expanding Learning: The learner learns the knowledge, which is distributed in the same layer, to broaden his horizon. The mechanism of searching learning material is to find the knowledge whose location label contains the same layer index.

  • Top-down Learning: The learner learns the knowledge in a specific domain from basis to proficiency. The mechanism of searching learning material is to find the knowledge whose location label contains the same domain index.

  • Knowledge-centered Learning: Having learned a specific knowledge, the learner start to study other related knowledge in order to strengthen the comprehension of the centered-knowledge. The mechanism of searching learning material is to find the knowledge which have a better fuzzy correlation with the focused knowledge and whose location label can be in any domain.

Applying a type of learning strategy indicates a definitive knowledge retrieval path. Based on the location label and fuzzy relationship, the knowledge retrieval process is performed by inductive indexing approach and the nearest neighbor algorithm along with the researching path. Since the search direction and connection between knowledge have been determined, a series of educational materials will be selected out after traversal.

337.2.1.4 Module 4 Learning Platform

The users can learn knowledge audiovisually from Learning Platform. The learning process consists of reviewing and learning.

In this article, we suggest that a reviewing process being introduced. According to Ebbinghaus forgetting curve, people forget more than 60 % of their learning things after 1 day. It is of little significance when people learn the knowledge only once. Therefore, we bring in an interactive process during learning activities, namely, the feedback from users about how difficult they feel about the content and how confident they feel they had understood the knowledge. The knowledge with higher difficulty or with lower score of understanding is prior to be reviewed.

There are three ways of using the system. Users can determine how to learn by themselves:
  • Passive learning
    • Receive knowledge material on screen

    • Go along with the instruction till to the end

  • Positive learning
    • Knowledge Map: well-constructed knowledge links in a graph on which man can select whatever he wants to learn.

    • Search Engine: always ready to retrieve knowledge

  • Interactive learning
    • Scoring: the user marks a knowledge unit with a score in certain area, so that the system response to modify its operation

    • Notebook: the user takes note for the learned knowledge and writes it on the board, in order to recall the memories by reviewing in the next round of learning. In this way, the efficiency of learning improves obviously.

337.3 Application Process

As illustrated in Fig. 337.3, the flow chat shows the system application process. Firstly, the user must open an account and configures his profile information. After logging in, the system will check user’s learning history. If the latest learning history exists, a review process will be organized by the system and the user could choose to review or deny. Then the Strategy Picker forms an instructional learning plan and selects appropriate knowledge materials out according to the user’s initial settings. Consequently, the educational contents will be provided in front of the learner. However, the learner could choose to learn things positively, to browse the knowledge freely and study whatever he wants; or to follow the learning instruction step by step. In both of the two learning processes the learner could have interactive activities. These activities as feedback will be sent back to Strategy Picker in real-time. Thus, the learning material and instructional plan could always be updated and personalized until the user has finished the learning activities. At last, the learning history and feedbacks in this round of learning will be stored and be prepared to serve the next round of learning.
Fig.  337.3

Application process of the system

337.4 Result Analyses

At first, the users input their profile information into the system by Entrance, which is not illustrated in this paper. The options of user profile are presented in Table 337.1. For every user, the profile configuration can only load in the first time, this process is not necessary later. In this case, the name of the learner is Wang Jianguo, the age 29 years old and so on.

Having received the entered user profile, KBS will determine the learning strategy and provide the learner with appropriate materials intelligently, according to user’s group, age, job, previous experience and purpose of learning. In this case, Jiang Jianguo is forklift driver and has little knowledge about logistics. The only knowledge he was familiar with is the forklift truck. Therefore, the system decides to use the knowledge-centered learning strategy. We can judge from the Fig. 337.4 that the user has already learned once at last time. As we can see, the knowledge in learning history provided by KBS is appropriately assigned. For instance, the classification of material handling equipment is firstly recommended, then the details of counterbalanced forklift, reach truck and later so on. The learner chooses one option of the learning history, namely the typical transmission organ of counterbalanced forklift truck. It shows that Jiang has given an understanding score with 3.5 point to this knowledge during latest learning. It means that he has to review it in time after learning, so that he can get a better learning performance. On the right side, some provided learning materials are listed for this knowledge. On the bottom side, the latest learning notes for this knowledge are shown in the box. They are helpful for the learners to recall their memories and to improve their learning performance.
Fig. 337.4

Page of learning history

After reviewing, as shown in picture Fig. 337.4, we choose to enter the next page to start a new learning. as picture Fig. 337.5 depicts, in this round of learning, the KBS provides not only the learning materials of counterbalanced forklift, but also the knowledge of reach truck. The available materials are also listed on the right side which concludes summarized document and video. Besides, the user can feedback his learning feelings by noting or scoring on the bottom of learning platform page. What are learned in this round will be recorded and shown as reviewing contents in the next round of learning.
Fig. 337.5

Page of learning center

337.5 Conclusions

KESL works with several characteristics. Knowledge materials in KBS can be added, maintained and updated to adjust to changes. Meanwhile, the learning strategy is well calculated and can be optimized by users’ feedback continuously. Different from a traditional educational system, KESL has also a review process which could improve the efficiency and effectiveness of user’s learning. In addition, the capability of customizing the learning interface could provide the users with good learning experience and keep them from the impression of boring learning. Furthermore, a direct-viewing demonstration of learning outcome could deepen their understanding of their efforts.

KESL is a professional frame for the further development of the knowledge-based education system in the logistics and supply chain management. The new knowledge can be easily integrated. The next development steps could be the application for smart phones and smart TV.


  1. 1.
    Sajja, Akerkar (eds) (2010) Knowledge-based systems for development. In: Advanced knowledge based systems: Model, applications and research. TMRF e-Book, vol. 1, pp 1–11Google Scholar
  2. 2.
    Li BM, Xie SQ, Xu X (2011) Recent development of knowledge-based systems, methods and tools for one-of-a-kind production. Knowl-Based Syst 24:1108–1119CrossRefGoogle Scholar
  3. 3.
    Ma X (2009) Design of a flexible E-learning system for employee’s education in manufacturing industry based on knowledge management. In: IEEE international symposium on it in medicine and education doi:  10.1109/ITIME.2009.5236228
  4. 4.
    Bittencourt II et al (2009) A computational model for developing semantic web-based educational systems. Knowl Based Syst 22:302–315CrossRefGoogle Scholar
  5. 5.
    Peredo R, Canales A, Menchaca A, Peredo I (2011) Intelligent web-based education system for adaptive learning. Expert Syst Appl 38:14690–14702CrossRefGoogle Scholar
  6. 6.
    Chow HKH, Choy KL, Lee WB et al (2005) Design of a case-based logistics strategy system. Expert Syst Appl 29:272–290Google Scholar
  7. 7.
    Qing Y, Yang Y, Chen J (2004) Goal-oriented platform based on knowledge-point: a new model of distance education system. In: Proceedings of the 18th international conference on advanced information networking and application (AINA’04), vol 2, pp. 528–531Google Scholar
  8. 8.
    Huang XL (2011) Study of personalized e-learning system based on knowledge structural graph. Proc Eng 15:3366–3370CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  1. 1.CDHK of Tongji UniversityShanghaiChina

Personalised recommendations