p-ISSN: 2301-5373

e-ISSN: 2654-5101

Jurnal Elektronik Ilmu Komputer Udayana

Volume 8, No 3. February 2020

Design and Development of Poultry Disease Classification with Certainty Factor Method

I Gusti Agung Ngurah Panji Palgunaa1, Luh Gede Astutia2

aInformatics Department, Udayana University Bali, Indonesia

1panji_palguna@student.unud.ac.id

2lg.astuti@cs.unud.ac.id

Abstract

Expert systems in organizations aimed at adding value, increasing productivity and managerial areas that can draw conclusions quickly. Like with organizations that conduct livestock business that are very promising but necessary high vigilance against disease as well as highly poultry susceptible to various types of diseases caused by viruses or bacteria. To know the disease quickly made a system that is useful for detecting, so breeders can check their poultry without seeing a veterinarian for early detection. Permanent Veterinarian required for further treatment.

  • 1.    Introduction

Expert system is a computer program that contains knowledge from one or more human experts about a specific field. This type of program is the first time developed by researchers of artificial intelligence in the 1960s and 1970s and commercially implemented during the 1980s. The general form of expert system is a program created based on a set of rules that analyzes information (usually given by the user of a system) regarding a class of specific problems and analysis mathematical problem. Depending on the design, an expert system is also capable recommend a series of user actions to be able to implement corrections. This system utilizes the capability of reasoning to reach a conclusion.

Some time ago there was an outbreak of disease in poultry which is very disadvantage to livestock entrepreneurs because not a few of these birds must destroyed, which is caused by an infectious and dangerous disease. Poultry is a type livestock groups of birds that are used for meat or eggs, which consumers are increasing every day. Therefore, this business is an opportunity very good for business. However, in this business disease is one of the risks which is high and must always be faced. Anticipate to prevent and recognize symptoms dangerous diseases are very important. The process of recognizing quickly and precisely from attacks of this type of disease is very difficult because the symptoms are generally similar and appear same. However, there are usually a number of symptoms that are typical for each type of disease poultry, for example in chickens.

The main modules contained in the expert system are (1) Reception module knowledge (knowledge acquisition mode). This module is used to collect knowledge that the system will use to draw conclusions from a problem. This module is intended for experts to input their knowledge into system; 2) Consultation module. This module functions to collect information about the symptoms of the problem being faced by the user, then processed by the system. This module is intended for users to input problems encountered in the system; and 3) Module explanation (explanation mode). This module serves to explain the decisions taken by the system.

  • 2.    Research Methods

The research method used is the prototype method. Prototype is a method in developing systems that use approaches to make things the program quickly and gradually so that it can immediately be evaluated by the user. Prototype represent the product model to be built or simulate structure, functionality and operating system. In making prototypes we can apply UCD

(User Centered Design) that is suitable for IT (Technology Information) lay users. Characteristics of UCD (User-Centered Design), namely:

  • a.    Understanding the user and his needs.

  • b.  Focus on the user in the early stages of design and evaluate the design results.

  • c.  Identify, make documentation and agree on the purpose and purpose user experience.

  • d.    Repetition is almost certain. The designers never succeeded only in one process.

The concept of inference engine is done by using the production rule (if..then) the mechanism is through forward chaining and weight assessment using the Certainty model Factors (CFs). The concept of user intervention and dialogue are developed by making user friendly interface for easy filling of data and facts. Output which presented in the form of information on the confidence value of the type of disease diagnosed with attack, while the validity test of the diagnosis results is compared with experts (breeders). In addition, application development and development are used by PHP and MySQL as language tools in its construction.

Poultry are animals that have wings, two legs, and lay eggs classified as a family of birds (aves). The most developed poultry animal for livestock business is a type of chicken. Chicken livestock have a high profit and potential. In addition to meat and eggs that are always increasing in consumption, chicken is very fast turnaround his business. However, not a few losses experienced by breeders because chickens are vulnerable to exposure disease. Disease in chickens can be caused by viruses, bacteria, inner parasites, outer parasites, and fungi. Various types of diseases caused by bacteria and viruses in chickens (Rasyaf, 2009) are presented in Tables 1 and 2 below.

b⅛f⅝U ιn∣tk Alt InkulMf Euiiipak bau

Mcnvrd kcpuliKpufilun IMUf pauhnya <1 k⅞akkαπ di 141 LlAl

Dudukden^Mi sikap Mctrthuπ∣∙kuk

GirMbtien


Giantwv


√>u<ujr


Nap*» cepat MrtM ∣w<M Scmpi WinjlMi Kaki p in VAna Saynp mcn^pnnhinp

NapM cepat l⅞ι⅛iuk>.ι Ieiui Itaeiiviua

Kuali Un Ieiur jelek Mcwrcb kehi IMi-Iuyauari

Figure 1. Poultry disease caused by a virus


ItunrXi DprMW


Ieiir


EfgDntp Avndrofflc W



KamaPvmakil NamaIjitiB               Gcjala

SjJsu makan Hrfkurang Biilukusain dari

Rictigtanl Duk KcitfiHtan πganlui din Iulu berdiri Iampak Jcsu

StrtuiiM LcHhUx-Intauan

Xafxu makan Hrrkurang IUdin hunt*

Bulu kusam dan nκ∙nχtanl

PruduJhi Iriur mm⅛run

StctKici hβ*Mψurdv>h Muka pucat

Hcrw-Kmin

Prudukxi telur HieniJun

Kciiipak mala kemerahan Kritar nanah dari mata dan Kmj

Pembengkakan Jan ruurs Ji n mold

  • Figure 2. Poultry Diseases caused by Bacteria

Method of Certainty Factors (CFs)

CFs express confidence in events based on events or on expert judgment. There are several methods of using CFs to handle distrust in knowledge-based systems. One way to use 1.0 for absolute trust and 0 for definite errors. CFs are not probabilities, but introduce concepts of trust and distrust. CFs indicate size certainty of a fact.

CFs [h, e] = MB [h, e] -MD [h, e]

Information:

CFs [h, e] = Certainty factor

MB [h, e] = Size of trust or level of confidence in hypothesis h, if given evidence e (between 0 and 1)

MD [h, e] = Size of distrust or level of confidence in the hypothesis h, if given evidence (between 0 and 1)

  • 3.    Result and Discussion

System work flow diagram implemented as follows:

Figure 3. Flowchart of System

Figure 4. Weight of each symptom and its correlation with disease

As for the user interface section, a menu showing is displayed several types of symptoms can be chosen (Figure 3). Users can determine the symptoms found in poultry. This symptom input is a premise for reasoning will be conducted on a knowledge base with a production rule that has been constructed. On generally, expert systems are developed using the LISP or Prolog language (Tsai et al., 1994). However, along with the development of tools that can be used for expert system development, several tools are already widely available with diversity and convenience offered (winexsys, shell, and so on). Language construction tools used in this study is PHP which is intended for ease in dissemination through the website.

System Implementation

a. Admin page

The Admin page is managed by an Expert in charge of inputting symptoms, knowledge and illness data from observations obtained. Can add, change, and erase symptoms, knowledge, and illness.

Figure 5. Admin page Symptoms List

Figure 6. Admin page Knowledge List

HalamanAonm



Users Lrafcing



twcwfd


Uawwre


IlMfl


DyllKSpN >∣rt ju<∣pu.IfMPRtatf >3 VΓTtιhll½lιi MKHPKi


AVtfynalran


MnLa"


Wtoapai


Xλ,⅞ΠSπMzM>QMtfιl4t*2!a∕fK}aJΓiteat> TrJraqMlajKkfiiztfJ LAJi


Λ∙afΛ<^an


UyllMMfuatanttKTiAucMlAyMujQrFlRtKiMiaitoL-FIKiCMiTWnG;


Jltndimsa


>M 1 t⅞∣Z⅛⅛n⅛*u MaacvXrtfRiaaFf <* ∙*m⅛ JUj ’ FRVtolHtfUKJDMC


AWTUMftQMMAlCCOtf


MSrtCin


OMUlJMg


DySina *tobτeufUH∕>Hrtfito WtfctfMTwMc-ItJrtaBUreufvrtJQitf:


ItfrypiFflaiw


H*KM∣ga∣fιiφ4aιHMK4ftK


Tsry-Ttffvre


oAUU-y


DySi mhm^a MismjMrMfMKyMmarMia j CCnJtfyotm u rmpqhm


s∣∙qmu


Si<Urtto>ft∙∣9flrτ9MQuC∣Md


MnMMQTarfO


0B 00 if [P 00 0B 0≡ P i


Figure 7. Admin page Users List


Figure 8. Admin Page Disease List


b. Users Page

Farmers can choose symptoms on this page, by selecting symptoms, the system will guess what illnesses suffered by poultry, for further treatment if needed.

Selamat Datang di Klasifikasi Penyakit

Unggas

Uiuli PaosaUl Urajgawtiu Utulunca Mai aki!

Figure 9. Initial Display The User page displays the Login / Register for use the system


Panp f⅛4juruι


Sistem Pakar Penyakit Unggas Metode CF (Certainty Factor)


Figure 10. List of Symptoms that the User Must Choose

Figure 11. CF Calculation Process

Figure 12. CF Calculation Results

  • 4.    Conclusion

Design of Poultry Disease Classification has been built with the Certainty Method Factor. As for the web pages presented there are 2 types, namely admin pages and pages user (user). Admin can add / change knowledge base, symptoms and disease. Users can find out what diseases are suffered by poultry based symptoms obtained, the user can also know the process of calculating the value of certainty factors to get the disease suffered by poultry.

The advantages of implementing an expert system for disease diagnosis very much depends on the results of calculating the level of confidence in supporting the process inference (reasoning)

of data and facts stored in the knowledge base. Method certainty factors can provide accurate results from calculating weights for conclusions resulting diagnosis. The use of CFs is very easy to determine weight given, and calculated based on facts that appear as symptoms which need to be considered in this Cfs method is giving weight values to symptoms caused will affect the amount of conclusions obtained. Not closes the possibility for further development of the Cfs method with a combination of rules more complex so that the complexity of the diagnosis can give more results satisfying

  • 5.    Suggestion

The suggestion that I can convey is that this system can be developed for more poultry diseases, because there is an admin page that makes it easy for Experts to manage the knowledge base.

References

  • [1]    Desiani, Anita, and Muhammad Arhami. "Konsep kecerdasan buatan." Yogyakarta: Penerbit Andi (2006): 161-183.

  • [2]    Fanny, Rahmi Ras, Nelly Astuti Hasibuan, and Efori Buulolo. "Perancangan Sistem Pakar Diagnosa Penyakit Asidosis Tubulus Renalis Menggunakan Metode Certainty Factor Dengan Penulusuran Forward Chaining." Jurnal Media Informatika Budidarma 1.1 (2017).

  • [3]    Kristanto, Andri. "Kecerdasan buatan." Yogyakarta: Graha Ilmu (2004).

  • [4]    Kusuma, Dewi. "Artificial Intelligence Teknik dan Aplikasinya." Graha Ilmu, Yogyakarta (2003)

  • [5]    Lenat, Douglas B., and Randall Davis. "Knowledge-based systems in artificial intelligence." New York: McGrav-Hill. Nev (1982).

  • [6]    Rasyaf, Muhammad. Panduan beternak ayam pedaging. Niaga Swadaya, 2012.

  • [7]    Rahwan, Iyad, and Guillermo R. Simari, eds. Argumentation in artificial intelligence. Vol.

  • 47.    Heidelberg: Springer, 2009.

  • [8]    Russell, Stuart J., and Peter Norvig. Artificial intelligence: a modern approach. Malaysia; Pearson Education Limited,, 2016.

  • [9]    Schalkoff, Robert J. Artificial intelligence: an engineering approach. New York: McGraw-Hill, 1990.

314