2023-SINTA4-Expert System Diagnose Eye Diseases Using

International Journal Software Engineering and Computer Science (IJSECS) 3 (2), 2023, 79-87
Published Online August 2023 in IJSECS (http://www.journal.lembagakita.org/index.php/ijsecs)

P-ISSN : 2776-4869, E-ISSN : 2776-3242. DOI: https://doi.org/10.35870/ijsecs.v3i2.1464.
Expert System Diagnose Eye Diseases Using

Case-Based Reasoning Method with Jaccard 3W
Algorithm
M Ibnu Sait
Computer Science Study Program, Faculty of Science and Technology, Universitas Islam Negeri Sumatera Utara, Deli
Serdang Regency, North Sumatra Province, Indonesia.
Email: [email protected]
Raissa Amanda Putri *

Computer Science Study Program, Faculty of Science and Technology, Universitas Islam Negeri Sumatera Utara, Deli
Serdang Regency, North Sumatra Province, Indonesia.
Email: [email protected]
Received: 4 July 2023; Accepted: 28 July 2023; Published: 13 August 2023.
Abstract: The Case Based Reasoning method is a method that applies 4 stages of the process, namely Retrieve, Reuse,
Revise, and Retain. The workings of the system in general are guided by the knowledge base possessed by the system
which originates from cases that have been handled by an expert. Then the level of similarity is calculated with the new
cases entered by the user using the 3W Jaccard Algorithm, based on the level of similarity of these cases, the system will
issue the conclusions drawn. This research resulted in an expert system application to detect eye disease. The method
used in calculating the Case Based Reasoning (CBR) method with the 3W Jaccard Algorithm. This application diagnosis
types of eye disease based on the results of the largest percentage and can provide a solution to the eye disease you are
experiencing.
Keywords: Expert System; CBR; 3W Jaccard; Eye.
1. Introduction
Every year in Indonesia, the number of people suffering from eye diseases is constantly increasing. Some of these
diseases include cataracts, blindness, refractive errors, and corneal problems. Eye diseases need to be examined and
treated by an ophthalmologist to prevent more serious damage. Eye disease is a disease characterized by disturbances and
abnormalities in human vision. A survey conducted by the Indonesian Association of Optometrists (PERDAMI) and
Litbangkes Agency of the population over 50 years old from 2014 to 2016 in 15 provinces found a blindness rate of 3%
[1]. This blindness is caused by cataracts accounting for 70.80%. The rate of visual impairment due to refractive error is
10.15%. However, the lack of knowledge about the many symptoms of eye diseases and limited information about eye
diseases, as well as the financial problems associated with the cost per consultation with the doctor. ophthalmologist,
leaving affected individuals unable to receive appropriate treatment.
To improve the treatment of eye diseases, much research and application development of web-based expert systems
has been carried out using Case-Based Inference (CBR) method [3][11][12][15]. This method involves using knowledge
of previous cases to diagnose new cases. Several studies have used this method to diagnose various diseases, including
eye diseases such as stye, dry eye, episcleritis, pterygium, and cataracts [3].
One of the related studies is the one conducted by Natalia in 2019, which used Bayes theorem to develop an expert
system for diagnosing eye diseases [1]. However, there is a difference in the approach compared to previous research, in
which the authors have built an expert system application to diagnose eye diseases using Case Based Reasoning (CBR)
method based on on the site [3]. This application is expected to improve the efficiency of medical services and reduce the
damage caused by the symptoms of eye disease, so that the patient can identify the type of eye disease easily and quickly.
quickly without having to see a doctor in person. With such an effort, it is hoped to raise public awareness of the
importance of early examination and treatment of eye diseases, as well as help those who need to be examined for vision-
related diseases to be diagnosed and treated. easier access. Therefore, the authors have built an expert system to diagnose
eye diseases by using case-based inference (CBR) based on the website to find out the differences compared to the
previous method. With this application, it can improve the performance of medical services and reduce the danger rate
Copyright © 2023 IJSECS International Journal Software Engineering and Computer Science (IJSECS), 3 (2) 2023, 78-87
79
80 Expert System Diagnose Eye Diseases Using Case-Based Reasoning Method with Jaccard 3W Algorithm
caused by disease symptoms, so that every patient with eye disease can easily and quickly find out what type of eye
disease is. eye disease without first visiting a doctor.
2. Research Method
The design of this study is to create a planning diagram of the lesion detection expert system by CBR case-based
inference with Jaccard 3w algorithm to detect eye diseases as an extension of the analysis process. demand accumulation.
This design is intended to provide an overview and design related to the digital system to be developed. The process of
conducting research applying Case Based Reasoning (CBR) method with Jaccard Algorithm 3W machine in detecting
eye diseases, specifically through several steps as follows:
Figure 1. System Flowchart
2.1 Case Based Reasoning

Case-based reasoning (CBR) is a way to solve new problems by reusing existing best-in-class knowledge, then performing
the process of adapting that knowledge. with new problems. Case-based inference (CBR) is a method used to implement
computer diagnostic systems in real-world applications. The concept of case-based reasoning arose from the idea of using
recorded experiences to solve new problems. Decision makers mainly use past problem-solving experiences to solve the
problems they are currently facing. CBR can also be used to analyze an issue based on the case in question and to further
categorize the case based on past classification experience. The advantage of RBC is that it allows us to use past case
examples to gain knowledge and ultimately know the topic. In Case Based Reasoning, there are four stages in the form
of retrieval, reuse, modification, preservation [2][6][7][11][13][16].
Figure 2. Case Based Reasoning (CBR) Cycle
2.2 Jaccard 3W Similarity Algorithm

Jaccard's 3W algorithm was first developed by Paul Jaccard from Switzerland. In this experiment, the 3W-Jaccard binary
algorithm is used to measure the similarity between two RDF (Resource Description Framework) nodes based on the
corresponding feature set. Intuitively, the more features two nodes have, the more similar they are. Let x be the size of
the feature set that belongs to nodes a and b, y is the size of the feature set that exists only in node A, and z is the size of
the feature set that exists only in node B. The 3W-Jaccard similarity is defined to give a higher weight on common features
and lower weight on distinguishing features, i.e., features that only exist at node A or B [5]. In the Weighted Jaccard 3W
Algorithm, similarity settings are made by giving a weight value to each symptom. This score is 5 (five) for severe
symptoms, 3 (three) for moderate symptoms, and 1 (one) for mild symptoms. The system will display the 5 (five) highest
value weights from the detection results made by the user. If the results of the similarity calculation are less than 0.6 (zero
point six), the user can provide suggestions and input. The data will go to the admin review page to find a solution. The
following is the 3W-Jaccard algorithm formula:
3. 𝑎
3W − Jaccard =
3. 𝑎 + 𝑏 + 𝑐
Information:
3 = Numbers for multiplication
α = the similarity value is 1 (new case) and 1 (old case)
b = the similarity value is 1 (new case) and 0 (old case)
c = the similarity value is 0 (new case) and 1 (old case)
Parameters weighting (w):

Severe Symptoms = 5
Moderate Symptoms = 3
Mild Symptoms = 1.
3. Result and Discussion
3.1 Results
3.1.1 Retrieve Process
To run an expert system for diagnosing eye diseases using the Case Based Reasoning method with the 3W-Jaccard
Algorithm, we need to enter the localhost/SPKR address, so the program can be used to make it easier for users to
determine eye disease based on symptoms and characteristics of eye disease without must ask directly to experts. input
by users with facts that exist in the knowledge base. Each symptom has a different weight value. Then the system will
carry out the calculation process by entering it into the 3W-Jaccard Algorithm formula. The process retrieve table can be
seen in table 1 below.
Figure 3. Retrieve Process
There is symptom data that has been previously entered by the admin on the system which is used as a knowledge base
consisting of code, symptom name, and symptom weight. The symptoms of eye disease along with their weight values
are as follows:
Table 1. Symptoms and Characteristics of Eye Disease

Code Eye symptoms Weight
G01 Red eye 3
G02 Dry eyes 3
G03 Watery eyes 1
G04 Blurry, blurry, and ghostly vision 5
G05 Excessive tears 3
G06 Swelling in the eye area 1
G07 Night blindness (Difficulty seeing at night or in dark places) 5
G08 Headache 3
G09 Sensitive to light 3
G10 Spots or lines appear on the eyes 5
G11 convulsions 5
G12 Daydream 1
G13 The eyelids cannot be closed 1
G14 Double vision 3
G15 Vomit 1
G16 Fever 1
G17 Sensitive to touch 1
G18 Great drowsiness 3
G19 Feeling unwell 1
G20 Weakening of the eye muscles 5
As for some types of diseases that are often experienced by the eye are as follows:
Table 2. Symptoms and Characteristics of Eye Disease

No Eye Disease Code
1 Cataract A01
2 Glaucoma A02
3 Conjunctivitis A03
4 Refractive error A04
5 Retinal Disorders A05
6 Amblyopia A06
The process of calculating the symptom data that is looking for uses the 3w-Jaccard algorithm formula to find the
similarity value to the symptom data from the previous damage:
3. 𝑎
3𝑊 − 𝐽𝑎𝑐𝑐𝑎𝑟𝑑 =
3. 𝑎 + 𝑏 + 𝑐
3.13
3𝑊 − 𝐽𝑎𝑐𝑐𝑎𝑟𝑑 = =
3.13 + 10 + 1
39
3𝑊 − 𝐽𝑎𝑐𝑐𝑎𝑟𝑑 =
50
3𝑊 − 𝐽𝑎𝑐𝑐𝑎𝑟𝑑 = 0.78
The calculation process for finding similarity values with the 3w-jaccard algorithm is displayed with 6 (six) other
diseases that will appear in the calculation results.
Table 3. Disease Similarity Value

Damage ID Disease Name Results
A04 Refractive Abnormalities 78%
A05 Retinal Disorders 75%
A01 Cataract 73.584905660377%
A06 Amblyopia 65.217391304348%
A02 Glaucoma 60%
A03 Conjunctivitis 41.666666666667%
Based on investigations from previous cases, the new case is like the 6 cases with the highest similarity value, namely the
type of refractive error with a value of 0.78 or around 78%.
3.1.2 Show program

The system was developed by implementing the PHP programming language and the MySQL database with XAPP and
Visual Studio Code as a text editor with the results of the implementation. The screen of the main page includes several
menus, namely the login menu for the administrator, the advice menu and the information.
Figure 4. Main Page Display
The appearance of the admin login page when you want to enter the administrator page. Admin can enter a username and
password to enter the administrator page.
Figure 5. Admin Login Display
The administrator's main page display consists of several menus such as types of disease, symptoms, rules, and
consultations. On this page the admin can make changes to diseases, symptoms, and rules such as adding new types of
diseases or changing symptoms of eye diseases.
Figure 6. Administrator page display
Consultation page display, on this page people with eye disease can directly consult with the system regarding damage
experienced by eye disease by directly selecting the symptoms experienced then just pressing the process button at the
very bottom which will then immediately display the results.
Figure 7. Display of the Consultation Page
The results page display is used to display the results of diseases experienced by the eye based on the symptoms that have
been previously selected.
Figure 8. Display of the Consultation Results Page.
3.2 Discussion
In this section, the results of the implementation of an expert system will be presented in diagnosing eye diseases by
utilizing the Case Based Reasoning method in collaboration with the 3W-Jaccard Algorithm. In the retrieve process stage,
this system facilitates the user to make a diagnosis of eye disease based on the symptoms and characteristics of the disease
entered, without the need for direct consultation with a medical expert. Each symptom is assigned a different weight, and
the system performs the calculation by applying the 3W-Jaccard Algorithm formula to measure the similarity between
the input symptoms and the previous eye disease symptom data. The process of calculating similarity using the 3W-
Jaccard Algorithm produces a similarity score between the entered disease symptoms and documented eye disease
symptom data. Through the results of this calculation, the system displays a similarity score expressed in the form of a
percentage. For example, in a new case situation with the input symptoms, the highest similarity was found with the type
of refractive error of 78%. In terms of program appearance, the implementation of this system uses the PHP Programming
Language and MySQL database with XAMPP as the local server and Visual Studio Code as a text editor. The main
interface of the system is designed with several menu options, including the login menu for the admin, the consultation
menu, and the information menu. Admin could make modifications to the types of diseases, symptoms, and regulations
through the administrator page. On the consultation page, users can directly consult their complaints by selecting the
symptoms they are experiencing and seeing the resulting diagnostic results. Diagnostic results are displayed through a
special page that displays results based on symptoms that have been previously selected. Through this implementation,
the expert system for diagnosing eye diseases has succeeded in providing solutions for users to diagnose eye diseases
efficiently and quickly based on the symptoms they are feeling, without having to consult directly with medical personnel.
The application of the Case Based Reasoning method and the 3W-Jaccard Algorithm in this system helps in identifying
potential types of eye diseases that may be experienced by users.
4. Related Work
In the domain of computer-assisted medical diagnosis, various approaches and methodologies have been explored to
improve accuracy and efficiency in identifying diseases. One particularly relevant approach is Case Based Reasoning
(CBR), which draws on past experiences to tackle new problems. In CBR, relevant knowledge from previous cases can
be reused and adapted to new situations. This section will provide an overview of related works in the field of medical
diagnosis and the application of CBR techniques.
Taufik Rachman (2021) developed an expert system based on forward chaining to diagnose eye diseases. This study
focuses on automatic reasoning processes to infer potential disease from observed symptoms. This approach enables early
detection and timely intervention, which contributes to better patient care [4]. Edwin Febriansyah (2021) designed a Case
Based Reasoning (CBR) system to detect problems with the Kawasaki KLX150 motorcycle. This study uses the 3W-
Jaccard similarity algorithm to evaluate the similarity between cases, which facilitates efficient problem solving and
decision making in vehicle maintenance [5]. Pratiwi (2013) presents the Case Based Reasoning (CBR) methodology for
diagnosing vehicle disorders using the Visual Basic programming language. This system uses past cases to diagnose
current problems, which speeds up the repair process and reduces downtime [8].
Jason Aldrian (2022) developed a backward chaining method combined with a certainty factor to diagnose problems
with Toyota Innova vehicles. This approach uses rule-based reasoning to infer potential problems based on observed
symptoms [9]. Ramadhana and Nababan (2021) used the Certainty Factor technique in the Case-Based Reasoning (CBR)
framework to diagnose eye disease in humans. This approach allows a more detailed assessment of symptom relevance
and provides insight into potential diagnoses [10]. This research, which was conducted by Umar (2023), introduced a
web-based expert system for the initial diagnosis of eye disease by combining the Forward Chaining and Certainty Factor
methods. The purpose of this system is to assist users in identifying potential eye conditions based on the symptoms
entered, so that medical attention can be provided in a timely manner [1]. Norma Jaya Telaumbanua (2022) proposes a
Case-Based Reasoning (CBR) system to detect eye disease. This study emphasizes the use of past cases to assist in
diagnosing current problems, indicating the potential of CBR in medical applications [3]. Tarigan (2023) developed a
Case-Based Reasoning (CBR) system for diagnosing disease, which shows the flexibility of the CBR approach in various
medical domains [6]. Yusmawati (2021) introduced the Case-Based Reasoning (CBR) method for diagnosing diseases in
rice plants, which highlighted CBR's adaptability in various contexts [7].
Syahfitri and Hartono (2022) explored the application of Pair Comparison and 3W-Jaccard similarities in diagnosing
problems in air conditioning systems. This research shows the importance of similarity assessment in identifying problems
[11]. Research conducted by I. Pratama, S. Wibisono, and E. Nurraharjo (2019) utilized the 3W-Jaccard similarity in the
Case-Based Reasoning (CBR) system to detect early neurological symptoms, which indicates the potential of this
algorithm in various diagnostic scenarios [12]. R. R. Ginting, U. Fatimah, and S. Sitorus (2020) developed an expert
system based on Dempster Shafer (DS) to detect problems with camera lenses, which shows the variety of techniques
available for specific diagnostic applications [13]. M. Brilliant (2022) implemented a Case-Based Reasoning (CBR)
system to detect stunting in children, which confirms the flexibility of the CBR technique in various health domains [15].
All the mentioned studies collectively demonstrated the effectiveness of using various reasoning techniques, such as
forward chaining, backward chaining, certainty factor, and 3W-Jaccard similarity, within a Case-Based Reasoning (CBR)
framework for medical diagnosis. This methodology proves the potential for increasing accuracy, efficiency, and speed
in diagnosing various health problems, including eye diseases. By leveraging experience and commonality assessments,
these systems provide valuable insights and support decision-making for medical professionals and patients. This research
also contributes to applying the Case-Based Reasoning method with the 3W-Jaccard Algorithm in the diagnosis of eye
diseases, through the application of this technique to an expert system designed.
5. Conclusion
Based on the results and discussion that have been produced in this study, it can be concluded that an expert system
application has been successfully developed to detect eye disease. The method implemented in the calculation is Case
Based Reasoning (CBR) with the 3W-Jaccard Algorithm. This application can diagnose types of eye disease based on the
results of the highest percentage of similarities and provide solutions to eye diseases that may be experienced by users.
The expert system produced in this study was designed using HTML and PHP programming languages, with data
management using the MySQL database. Further development related to this application can involve the ideas that have
been generated from this research, to continue to improve the quality and effectiveness of the system in providing eye
disease diagnoses.
Acknowledgments
The author would like to thank all the reviewers who have contributed to the process of peer review of manuscripts in
this matter. The professional support and assistance of all respected reviewers has made this journal eligible for
publication. As well as lecturers, parents and friends who also provided support in writing this research.
References
[1] Umar, F., 2023. Sistem Pakar Berbasis Web untuk Diagnosa Awal Penyakit Mata dengan Penerapan Forward
Chaining dan Certainty Factor. Jurnal Ilmiah ILKOMINFO-Ilmu Komputer & Informatika, 6(1), pp.54-66. DOI:
https://doi.org/10.47324/ilkominfo.v6i1.166.
[2] Syahfitri, D.N. and Hartono, B., 2022. Pembobotan Atribut menggunakan Pairwise Comparison pada CBR Deteksi
Kerusakan AC dengan Algoritma Similaritas 3W-Jaccard. Explore IT!: Jurnal Keilmuan dan Aplikasi Teknik
Informatika, 14(1), pp.1-8. DOI: https://doi.org/10.35891/explorit.v14i1.2938.
[3] Telambanua, N.J., Nofriadi, N. and Dermawan, A., 2022. Sistem Pakar Untuk Mendeteksi Penyakit Mata
Menerapkan Metode Case Based Reasoning. Building of Informatics, Technology and Science (BITS), 4(2),
pp.570-580. DOI: https://doi.org/10.47065/bits.v4i2.2116.
[4] Rachman, T. and Pratiwi, R.S., 2021. SISTEM PAKAR DIAGNOSA PENYAKIT MATA MENGGUNAKAN
METODE FORWARD CHAINING (Studi kasus RUMAH SAKIT UMUM DAERAH LAWANG). Jurnal
Mnemonic, 4(2), pp.38-44. DOI: https://doi.org/10.36040/mnemonic.v4i2.4123.
[5] Febriansyah, E. and Winarno, E., 2021. S SISTEM PAKAR DIAGNOSA KERUSAKAN MOTOR KAWASAKI
KLX150 MENGGUNAKAN METODE CASE-BASED REASONING DENGAN ALGORITMA 3W-
JACCARD. Jurnal Ilmiah Informatika, 9(02), pp.68-74. DOI: https://doi.org/10.33884/jif.v9i02.3747.
[6] Tarigan, P., 2022. Sistem Pakar Metode Case Based Reasoning Mendiagnosa Penyakit Kanker
Tenggorokan. KAKIFIKOM (Kumpulan Artikel Karya Ilmiah Fakultas Ilmu Komputer), pp.100-107.
[7] Yusmawati, Y. and Sanusi, S., 2021. Sistem Pakar Diagnosa Penyakit Pada Tanaman Padi Menggunakan Metode
Case Based Reasoning (CBR) Berbasis Website. Jurnal Real Riset, 3(2), pp.187-192. DOI:
https://doi.org/10.47647/jrr.v3i2.425.
[8] Pratiwi, H., 2013. Sistem Pakar Diagnosa Kerusakan Mobil Menggunakan Metode Case Based Reasoning
Berbasis Visual Basic. Digilib. Mercubuana. Ac. Id, 5(2).
[9] Aldrian, J. and Setiabudi, D.H., 2022. Sistem Pakar Diagnosa Kerusakan Pada Mobil Toyota Innova Dengan
Metode Backward Chaining Dan Certainity Factor. Jurnal Infra, 10(2), pp.42-48.
[10] Ramadhana, F. and Nababan, A.A., 2021. Sistem Pakar Diagnosa Penyakit Mata Pada Manusia Menggunakan
Metode Certainty Factor. Jurnal Sistem Informasi Dan Teknologi Jaringan (Sisfotekjar), 2(2), pp.36-40.
[12] Pratama, I., Wibisono, S. and Nurraharjo, E., 2019. Sistem Pakar Deteksi Dini Gejala Gangguan Neurologi
Menggunakan Metode Case Based Reasoning dengan Algoritma Similaritas 3W-Jaccard. 3(3), pp. 443–449.
[13] Ginting, R.R., Saripurna, D. and Sitorus, U.F.S., 2022. Sistem PakarMendeteksi Kerusakan Lensa Cannon 1885
Pada Camera Cannon 1200 Dengan Mengggunakan Metode Dempster Shafer (DS). Jurnal Cyber Tech, 4(4). OI:
https://doi.org/10.53513/jct.v4i4.1887.
[14] Ginting, T.O.B., Saripurna, D. and Pranata, A., 2022. Sistem Pakar Mendiagnosa Penyakit Abortus Menggunakan
Metode Certainty Factor. Jurnal Cyber Tech, 2(10). DOI: https://doi.org/10.53513/jct.v2i10.734.
[15] Brilliant, M. and Nizamiyati, N., 2022. Sistem Pakar Metode Case-Based Reasoning Untuk Deteksi Penyakit
Stunting Pada Anak. Jurnal SIMADA (Sistem Informasi dan Manajemen Basis Data), 5(2), pp.13-22. DOI:
https://doi.org/10.30873/simada.v5i2.3415.

2023-SINTA4-Expert System Diagnose Eye Diseases Using

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2023-SINTA4-Expert System Diagnose Eye Diseases Using

Uploaded by

Copyright:

Available Formats

International Journal Software Engineering and Computer Science (IJSECS) 3 (2), 2023, 79-87

Published Online August 2023 in IJSECS (http://www.journal.lembagakita.org/index.php/ijsecs)

Expert System Diagnose Eye Diseases Using

Raissa Amanda Putri *

Received: 4 July 2023; Accepted: 28 July 2023; Published: 13 August 2023.

Keywords: Expert System; CBR; 3W Jaccard; Eye.

Figure 1. System Flowchart

2.1 Case Based Reasoning

Figure 2. Case Based Reasoning (CBR) Cycle

2.2 Jaccard 3W Similarity Algorithm

Parameters weighting (w):

3. Result and Discussion

Figure 3. Retrieve Process

Table 1. Symptoms and Characteristics of Eye Disease

Table 2. Symptoms and Characteristics of Eye Disease

Table 3. Disease Similarity Value

3.1.2 Show program

Figure 4. Main Page Display

Figure 5. Admin Login Display

Figure 6. Administrator page display

Figure 7. Display of the Consultation Page

Figure 8. Display of the Consultation Results Page.

You might also like