Expert System for Areca Plant Disease Detection Using Forward Chaining Method

Areca palm has been widely planted by the people of the Asahan Regency. This plant is cultivated not only by the farmers, but also by many people living in this regency. Therefore, an expert system is needed to help villagers in detecting areca palm desease. In areca palm cultivation, pests and diseases attack always happen. The purpose of this research is to apply an expert system as a tool to diagnose and also provide advice for control. This expert system, in the form of a consultation scenario, can be done by answering every question (yes or no); all answers are compared to the areca palm plants needs based-on expert experience. The inference method used is forward chaining. The output of this system is the types of


Introduction
Nowadays, data innovation in a gadget has an ability to cover a wide network. Frameworks in computers make it possible to store a lot of information in a short time, and make the computer can think. The computer development can help in solving problems in everyday life and in various fields.
One of the crops grown by local farmers beside oil palm is areca nut. Based on the data from the Agriculture Office of Asahan Regency, areca nut plants are also easy to find. The development of these plants is not complicated, hence, anyone who is not a farmer can cultivate these plants easily (DISBUN, 2017). Areca nut cultivation faces various obstacles, especially pests and diseases that can cause a decrease in yield (quality and quantity).
This research proposes an application to check a desease on areca plant using an expert system. Therefore, it can be used by many villagers. Forward chaining was used as a main method. The output of this expert system is to show data about side effects and infections that occur in areca nuts and how to control them.

Literature Study
Expert System is a framework that receives human information and inserts it into a computer. Therefore, it has a capacity to think critically like a specialist. With this Expert System, ordinary people can solve their problems or just look for quality data that must be obtained with the help of specialists in their fields (Irawan & Nasution, 2018). Expert systems are used to unravel problems that are difficult to unravel by utilizing common projects, given the costs involved in creating specialist frameworks that are far more prominent than standard frameworks (Ritonga & Irawan, 2017).
Another study shows that expert system is a visible framework for receiving human information into a computer, with the aim that the computer can handle problems like specialists Jurnal Piksel 8(2): 115 -122 (September 2020) in solving the problem. The essence of the clinical master's framework is to assist specialist analytical procedures. It thinks about reality and its side effects to give it determination. It concludes that the clinical master framework uses information about disease and reality about patients to recommend findings (Munaiseche, Kaparang, & Rompas, 2018). With this main framework, application administrators can explain a confusing problem that is generally understood by a specialist (Rukun et al., 2017). This Expert System works by utilizing information from the specialists according to their area of expertise. Expert system consists of two principle segments, i.e. the information base and the ultimate goal of making device information obtained from gathering information in a particular field. Information in this context is characterized as a variety of information and many rules for controlling or handling information to turn into new information. The information base is an important part of an expert system, the capacity of a specialist framework is usually determined by the limits of its information base, while a dynamic engine is an application that assists and assists the client framework in controlling information and selecting the appropriate information to get results (Hawa, Abdullah, & Usman, 2015).
Expert systems consist of domain experts, designers, inference engines, knowledge base, user interface, and users. There are relationships between these components. Domain experts connect to the knowledge base to provide rules and facts. Domain experts are usually experts in a field or field. The knowledge base stores collected rules and facts. The knowledge base is also connected to an inference engine which is used to process rules to infer a set of rules or facts. Inference engines are usually designed by programmers or designers. The inference engine is then connected to the user interface which is used to collect data from the user. It is also developed by the designer. This trend can also be followed backwards. The user interface provides information to the inference engine and knowledge base for user data to be processed. Also for knowledge base updates, the need to contact a domain expert is required. All of this can be represented in Figure   In expert system study, several investigative procedures were found, one of which is forward chaining. The forward chaining method can be used to check for certain diseases, e.g. humans, creatures, and plant infections (Sinaga et al., 2019). Nowadays, there have been several developments in the results of expert systems in various fields according to one's expertise. One example of implementing an expert system in the health sector is the application of eye disorders diagnosis and eye care tips with the web-based forward chaining method (Pramesti, Arifudin, & Sugiharti, 2016).
Forward chaining is a tracing process that begins by displaying a compelling collection of data or facts leading to a final conclusion. Therefore, the forward chaining method starts from first input information (if section) and then to conclusions (then section) or in a model IF (Enter Information) and THEN (conclusion). Input information can be data, evidence, findings or knowledge, while conclusions may be interesting, explanation, or diagnosis (Rukun et al., 2017).
The forward chaining design phases are symptom data, disease data, decision tree, and inference engine (Shofi, Wardhani, & Anisa, 2016). Another phases in the research method on forward chaining are determining initial facts, determining rules, determining decision trees, and acquiring knowledge (Samsudin, 2018), (Riyadi & Samsudin, 2016). The forward chaining method has several characteristics, namely: (1) Planning, monitoring and control (2) Presenting for the future (3) antecedents for consequences (4) Data guidance, bottom-up reasoning (5) Hoping to get what solutions follow the facts (6) The extent of the first search is easier (7) antecedents determine the search (8) Explanation is not facilitated (Arhami, 2005).
Research on expert systems, especially using the forward chaning method in agriculture, has been carried out in the diagnosis of disease in cocoa (Hawa et al., 2015). Another implementation in the field of human diseases using forward chaining method was an expert system for diagnosing gastric disease (Samsudin, 2018).
Similar research on areca nut has been carried out by Salim with several diseases and pests in areca nut, namely Brontispa longissima, Tirathaba, Aphids, Male Flower Borer, Grasshopper (Valanga nigricornis), Leaf blight disease (Salim, 2012). The results showed the diseases and pests in areca nut can be identified. Therefore, in this study, an expert system is applied using the forward chaining method. The expert system proposed in this study aims for diagnosing areca nut disease will be implemented on a website. The website has many advantages, such as a multimedia information, among others. The process is carried out using a web browser such as Mozilla Firefox and Google Chrome, etc. (Irawan & Simargolang, 2018).

Reseach Method
This research used several steps as shown in Figure 2. The data was collected from references related to research topics, including journals, e-books, etc. The problem is identified as something that hinders the research objectives. Problems must be followed up to find a solution. This will need system requirements analysis, which identifies the requirements needed in making an application for diagnosing areca nut disease. System requirements analysis includes areca nut disease data.

Data Collection and Problem Identification
Rule-based Generation Inference Machine Building User Interface Design The formation of a knowledge base was carried out after the data was collected. The formation of a knowledge base is classified with a code on each data. Several data is needed in thus study, i.e. symptoms of disease data, disease data, and solutions data. Furthermore, the formation of an inference machine according to the forward chaining method. And the last is the design of the user interface which is a display of communication to the user.

Data Collection and Problem Identification
Data collection is obtained from the results of agricultural research which is then identified and analyzed. Additional data results were from interviews to farmers and the Asahan District Agriculture Office staffs. The results are symptom data, areca nut disease and solutions. Jurnal Piksel 8(2): 115 -122 (September 2020)

Rule-Based Generation
In the concept of an expert system for examining diseases of areca plants, it is very necessary to find the symptoms using the forward chaining method. Examination of diseases in areca plants will be grouped into symptoms of disease, areca plant disease, and data solution.

Symptoms of Disease
The knowledge base of areca plant diseases can be seen Table 1.

Data Solution
Solution data from areca plant disease can be seen in Table 3.  Figure 3 shows the formation of the inference engine. Search engine was prepared for solutions to areca plant disease with the symptoms of areca nut disease (Figure 3). After the inference engine is formed the production rules are prepared as shown in Table 4.  Solution  1  IF GP01 THEN PP01  SPP01  2  IF GP02 THEN PP02  SPP02  3  IF GP03 AND GP04 THEN PP03  SPP03  4  IF GP04 THEN PP04  SPP04  5  IF GP11 AND GP12 AND GP13 THEN PP05  SPP05  6  IF GP09 AND GP10 AND GP11 AND GP12 AND GP13 THEN PP06  SPP06  7  IF GP06 AND GP07 AND GP08 THEN PP07  SPP07  8  IF GP14 AND GP15 AND GP16 AND GP17 THEN PP08  SPP08  9  IF GP18 THEN PP09  SPP09  10 IF GP20 AND GP21 AND GP22 THEN PP10 SPP10 11 IF GP23 AND GP24 AND GP25 AND GP26 THEN PP11 SPP11 12 IF GP27 THEN PP12 SPP12 13 IF GP28 AND GP29 THEN PP13 SPP13

User Interface Design
After the knowledge base and inference engine have been established, the next step is designing a web-based user interface. The diagnostic form shows an important role in determining the outcome. In the diagnostic form, there are questions related to disease symptoms using the forward chaining method. The diagnostic results show diagnostic data information, types of diseases, general symptoms, diseases, and treatment and prevention.

Conclusion
The study has been implemented a forward chain-based expert system. The result showed the inference engine has been successfully predict a deasese, starting from the fact symptoms of areca nut collection and the areca nut disease prediction. Therefore, a solution for handling the disease can be taken which can become beneficial information to farmers for which the farmers can take good care of areca plants. This system is also useful for the Asahan Regency Agriculture Office in socializing how to properly care for areca plants by showing symptoms of the disease. So that not only the data of areca planters is high, but the areca nut production can also be balanced.