Ontological and Sensor-Based Supporting System on Damage Prevention for Farming: a Case Study of Phetchabun Tamarind Farms

Phimphan Thipphayasaeng1 and Sant Phanichsiti2 Phetchabun Rajabhat University, Phetchabun, Thailand [email protected] [email protected]

Received: 18/10/2019 Accepted: 27/11/2019

Abstract - To achieve products in quantity I. INTRODUCTION and quality, farmers need both knowledge on how to handle problems and precise A tamarind is an economic fruit famous for relevant farm data. Tamarind fruits are the its antioxidant and anti-inflammatory properties [1- major economic product of Phetchabun, 4]. There are several health benefits from Thailand, and they require a countermeasure consuming tamarind fruits including improving against insects and diseases to damage digestion [2-3] and strengthening liver and them. This work presents a combination of heart [5-6]. Moreover, they can be used in knowledge-based and sensor-based framework pharmacological usage as herbal ingredients to support tamarind farmers in handling for producing medicines for treating cough, invading diseases and insects. An ontology cold and asthma [7]. In Thailand, agriculture is designed to create a network of relevant of is mostly located in Northern and farming concepts for smart decision- North-eastern region, especially in Phetchabun, Loei making while sensors are exploited to and Uttaradit for most productivity rate obtain precise and real-time data of the respectively [8]. For Phetchabun, tamarind is farm climate. Together, they provide a the most important export fruit which values service to alert once farm circumstances about 1,200 million baht annually [8]. are likely to be invaded by pests and diseases. Furthermore, recommendation on Same to a cultivation of other fruits, how to handle the problem and information productivity and growth rate of tamarind is about tamarind in which are established in affected by diseases and pests, especially in a the ontology are accessible for farm caretakers. pre-harvesting phase [9-14]. The damaged In evaluations, the framework shows fruits are devalued and can rarely be sold. promising results from achieving 0.85, 0.97 Commonly, prevention of the damage is and 0.91 in average for precision, recall preferable, but applying too much pesticides or and f-measure, respectively. chemical compounds should be avoided. In fact, the invasion of disease and pests is Keywords - Ontology, Sensor-Based Supporting however able to be specifically prevented if System, Tamarind Farming realized sooner, but most of the cases are recovery from the invasion.

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 20 Phimphan Thipphayasaeng and Sant Phanichsiti

In this work, we propose a combination of II. LITERATURE REVIEWS knowledge-based system using ontology representation [15-17] and sensor detection for A. Tamarind Farming and an Invasive Insects supporting tamarind agriculture. The main and Diseases objectives include to alert farmers for possible Tamarind (Tamarindus indica) is a leguminous pest and disease invasion, to provide tree in Fabaceae family. The trees bear a fruit information for detecting of tamarind-related in a pod-like shape containing a brown pulp harmful organisms, and to suggest a method to which is edible and has several culinary uses prevent and cure such threats. Moreover, the around the world. The pulp is also used in knowledge providing in a developed ontology traditional medicine and as a metal polish. A is not only about commonly used agricultural wood from the trees can be used for method, but also organic method regarding woodworking and tamarind seed oil can be ‘Sufficiency Economy Farming’. This work extracted from the seeds. Tamarind is cultivated thus is expected to expedite tamarind farming around the world in tropical and subtropical dynamically from difference in individual zones. farm settings and location with data from employed sensors. The rest of this paper is The tamarind has been naturalized in South- organized as follows. Section 2 provides east Asia including Indonesia, , Sri Lanka, details of background knowledge and related Philippines, and Thailand while Thailand has works in smart farming and agriculture- the largest plantations among these nations, supporting applications in Thailand. Section 3 followed by Indonesia and Myanmar. Similar proposes our methodology for developing the to other fruits, diseases and pests are the main ontological and sensor-based supporting issues in tamarind cultivation. system on damage prevention for tamarind farming. In Section 4, assessment of the For invasive insects and diseases, there are developed ontology and the application is researches about timing and factors of their experimented and reported. Section 5 gives a invasion as exemplified in Table I. conclusion and discussion of the paper as well as the plan for future work.

TABLE I SHOW AN INVASIVE INSECTS AND DISEASES IN TAMARIND

Type Name Seasoning/Timing Damage Phase Insect Cockchafer (Microlrichai Sp.) April-June / night Bud and leaf Budding phase [11, 18-20] Lavae of Castor semi-looper Rainy Season / entire Leaf and fruit Fruiting phase (Achaea janat) [11, 18-19] day Citrus fruit borer (Citripestis December and Fruit pod Fruiting phase sagittiferella) [11, 18-19] January Scale insect (Diaspididae) All time Branch, bud, Growing, Budding, [11, 18-19] fruit pod Fruiting phase Bean weevil (Caryedon gonagra) All time Fruit pod Budding and [11, 18-19] Fruiting phase Disease Rotten pod disease (Phomopsis Rainy Season Fruit pod Budding and sp.) [9-10, 18-19, 21] Fruiting phase Powdery mildew (Pseudoidium Rainy and cold season Fruit pod Budding and tamarindi) [9-10, 18-19, 22-23] Fruiting phase

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 21 Ontological and Sensor-Based Supporting System on Damage Prevention for Tamarind Farming: a Case Study of Phetchabun Tamarind Farms TABLE II SHOW RESEARCH AND APPLICATION FOR SUPPORTING AGRICULTURE IN THAILAND

Technology / Title Main Task / Purpose User / Application type Type Technique [15] Developing an ontology from analysis of Product Knowledge Ontology Research / 40 web pages of a domain Representation, representation Method Information Search [24] The framework to provide a suggestion on Durian Individual-based Ontology Research / how to be self-reliant and sustainable farm farmers recommender representation Method system and inference [25] A sensor-based environment providing Thai Applications and Sensors and Research / real-time warnings of risks in farm farmers Services models for Method operations, such as disease alarms and weather and warning on approaching rain showers disease forecast [26] A geographic information system Farmer Ecosystem and Geographic Mobile application about Thai geography for Delivery information application Agriculture management system [27] A recommendation system on rice Thai Information Geographic Mobile cultivation planning by considering of farmers Provider information application geographic and climate information system and rule inference [28] A recommendation system for Thai farmer Thai rice Individual-based Information Mobile / on fertilization and soil quality regarding farmers recommender system and rule Web farm location system inference application [29] Web-based information provider in Thai Individual-based Knowledge Web Portal various aspects on rice cultivation farmers recommender organization system [30] A personalized recommendation system Thai rice Information Ontology Research / for Thai rice farmer on how to care for rice Farmers provider representation Method paddy and inference [31] Application with the Internet of Things Thai rice Individual-based Sensors for Mobile Technology Control in Smart Farms Farmers recommender environment application Mushroom system monitoring [32] IoT and agriculture data analysis for smart Thai Individual-based Sensors for Mobile farm mushroom recommender environment application farmers system monitoring [33] Using IoT to sense the change of Thai Individual-based Sensors for Mobile environment in the farm for precision farmer recommender environment application farming system monitoring [34] A recommendation framework for Thai Individual-based data-driven Web-based managing swine breeding farm farmer recommender model Information system management System

These organisms are an invasive being that B. Smart Farming and Agriculture-Supporting is harmful to the tamarind in common. Applications in Thailand Normally, the professional farmers know There are several works aiming to use about them and are prepared in handling them information technology for supporting agriculture in based on incident history of their farm. Thailand. They can be classified into two types However, if the outbreak occurs out of their as research and application. For research, method prepared time (from climate changes), the and model are proposed in publications with damage will be heavy and required urgent the latest technology and idea. On the other treatment. For treating the diseases, chemical hand, applications are those services accessible and medicines are commonly used. Specified usable from website, personal computer or chemical insecticides are applied to annihilate mobile applicant. We summarize them with the invasive pests. Though, these chemical brief details in Table II. substances can save the day, it is commendable to use them as less as possible; thus, prevention From the review, the core technology in of the outbreak is more preferable. these works can be categorized into two main

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 22 Phimphan Thipphayasaeng and Sant Phanichsiti groups which are information-based system and property (part-to-whole relationship). For and sensor-based system. The information- property relation, there are two subtypes as based systems focus on assisting to users in object properties (part-of relation; class to terms of decision-making, providing clear class) and data properties (attribute-of relation; value detail on handling problems and optimizing to class). their productivity. The most used information representation from these works is an ontology For covering all aspects needed for tamarind since it is claimed for its ability in sharing farming services, the ontology (translated to knowledge for both machine and human [35- English) as shown in Figure 1 is designed 36]. For sensor-based systems, detection of using HOZO ontology editor [42], and details climate and environment is their aim in which of some major classes are given in Table III. can help farmers in realizing precise factors in farming. In this work, we combine the use of The ontology is designed to represent both sensor and knowledge representation to network of information as knowledge representation cope with tamarind farming issues. We expect for tamarind farm management. The relations that the real time detection of sensors and have a coverage to cultivation in the farm and decision-making of an ontology with inference how to handle possible issues. The classes are can effectively help farmers to rationally solve scoped to only information related to tamarind their issues in time. farming; hence, other information such as scientific name, color of the leaf and shape of III. RESEARCH METHODOLOGY the tree is omitted in this ontology.

A. Design of Tamarind Management Ontology The ontology as knowledge representation An ontology [37-39] is chosen to create a though is unable to work in an application network of related information regarding without instances [40]. In this work, OAM tamarind farm management. In this ontology, [43] is chosen to assist in instantiation by information such as disease, pest, tamarind mapping a database schema to a schema of the tree morphology and environmental factors developed ontology. The instances in this along with individual farm information is work include the environment factors gaining theoretically connected into a network of from the sensors and the individual farm concepts. As ontological components [40-41], information in which is obtained from the user concepts are designed as classes, and the input. concepts are related to other by assigning relations including is a (hierarchical relationship)

Figure 1. Developing the Ontological and Sensor-Based Supporting System on Damage Prevention for Tamarind Farming

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 23 Ontological and Sensor-Based Supporting System on Damage Prevention for Tamarind Farming: a Case Study of Phetchabun Tamarind Farms

Figure 2. The Core Concepts of Tamarind Management Ontology

TABLE III DETAILS OF MAJOR CLASSES IN THE TAMARIND MANAGEMENT ONTOLOGY

Ontology Definition Hierarchical Relation Property Relation Class Harmful  Having 2 subclasses  Having 3 part-of property Organism o Disease o damaging part [Tamarind o Pest Morphology] o invading phase [Cultivation phase] o suitable environment factor [Environmental Factor]  Having 2 attribute-of property o name [string] o ID [string] Disease harmful diseases  Subclass of Harmful Organism  Having 1 part-of property specifically to  Having 3 subclasses o possible booming [Environmental tamarind trees o Fungal Disease Factor] o Viral Disease  Having 4 attribute-of property o Bacterial Disease o pathogen name [string] o symptom description [string] o prevention method ID [string] o cure method ID [string]  Having 5 Inherited properties from Harmful Organism Pest harmful animals  Subclass of Harmful Organism  Having 2 attribute-of property specifically to  Having 3 subclasses o prevention method ID [string] tamarind trees o Insect o cure method ID [string] o Bird  Having 5 Inherited properties from o Rodent Harmful Organism Tamarind a part of tamarind  Having 4 subclasses none Morphology tree based on o Transportation System functional o Absorption System systems o Photosynthesis System o Reproduction System

Environmental Factors about none  Having 2 attribute-of property Factor climatic o Humidity [integer] environment o Temperature [integer] o Daytime period [integer] o Rainfall volume [integer]

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 24 Phimphan Thipphayasaeng and Sant Phanichsiti

B. Services of the Proposed Framework of a property in the designed ontology while 1) Alert Service the ‘result part’ is the instances linked to the The alert service is designed to give a class in the Harmful Organism tree. In warning to farm caretakers about the possible representation, the ontological production rule invasion of disease and pest regarding is defined as follow. environmental factors. Since these harmful organisms have a specific set of factors to IF {condition1 and condition2 and … become booming, we decide to apply sensors conditionn } THEN {result} to detect environmental factors of the tamarind fields for the more precise alert. Thus, the alert The result part is triggered only if all the will be activated and informed to users when conditions are matched. In the condition part, the environmental factors are matched to the the assigned operation depends on the type of pre-defined rule set based on farming knowledge. ontological relation. For part-of relation and string data type of an attribute-of relation, A detail on the chosen sensor types to exact and partial string-matching is allowed. detect environmental factors is related to the The integer-based attribute-of relation however knowledge of tamarind farm management allows basic mathematic operations including about a chance for suitable environments of greater than, less then, equal to, in between to disease and pest booming. In this work, four match up property value. The rules to activate types of sensor as shown in Table IV are alerts are exemplified in Table V. selected for obtaining the different factors in a tamarind farm. The rules are obtained by extracting knowledge from publications and guidelines The outputs of the sensors are then stored from Department of Agricultural Extension, in a database which is mapped to the ontology. Thailand. For example, details of the ‘powdery These values are then sent to the ontological mildew disease’ are extracted from [9-10, 18- rule matching modules, and if matched, the 19, 22-23], and information of ‘cockchafer’ alert is triggered and informed to the farm (Holotrichia sp.) is from [11, 18-20]. The rules caretakers. The ontological rules in this work then are reviewed and approved by tamarind are the production rule in which contents of experts. the ‘condition part’ are the classes and values

TABLE IV EMPLOYED SENSORS AND THEIR DETAILS

Measuring Environmental Duration / Frequency Sensor Type Employing Area unit Factor of detection Non-contact temperature Celsius (ºc) Air temperature Hourly Located in a field sensor Non-contact humidity Percentage (%) Air relative humidity Hourly Located in a field sensor Sunlight sensor Minute Daytime period Continuously Located near the field Rain gauges and Millimetre Frequency and Daily Located near the precipitation Sensors amount of rainfall field

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 25 Ontological and Sensor-Based Supporting System on Damage Prevention for Tamarind Farming: a Case Study of Phetchabun Tamarind Farms TABLE V THE ONTOLOGICAL PRODUCTION RULE

Rule Condition part Result part 1 Humidity ≥80% & Rainfall ≥ 100 mm. pod rot disease (Phomopsis sp.) Humidity greater than or equal to 70% & Daytime ≥ 13 2 powdery mildew disease hours 3 Rainfall <30 mm. & average temperature at night <20 ºc Butterfly larvae average temperature at day time ≥ 33 ºc & Daytime > 12 4 cockchafer (Holotrichia sp.) hours

2) Information Providing Service knowledge engineering and domain knowledge. The This service provides information related to second experiment is to evaluate the results of harmful organisms. The information is a text- the framework. By comparing the Alert service with based content for users to read as a guideline a decision of farm caretakers, we can see that in tamarind farming. This service is expected the knowledge provided in the ontology is to be an additional knowledge provision to resemble to tacit knowledge of farm experts or users to be aware of harmful organisms that not. may possibly occur extraordinarily outside of their prevalent season. There are two functions A. Ontology Evaluation in this service: general information for browsing and In this evaluation, 5 ontology experts were specific information regarding alert. asked to assess the designed ontology and its rule set. They were asked to answer the For the former, users can browse descriptive questionnaire regarding quality of the ontology. The information based on ontology concepts by questionnaire is in Likert scale [44] in which 5 navigating through the network of the developed is ‘totally agree’ to 1 refers to totally disagree. ontology. The content in which is stored in There are 6 quality aspects as given in Table 6. attribute-of relation explains the details of the The evaluation result in average from 5 class. For instance, the symptom relation experts is shown in Figure 3. (attribute-of relation of the class ‘Disease’) of the ontology gives a detail of how the Ontology Evaluation Results symptom of the disease is and which part of a tree to be noticed. Coverage 5 4 Correctnes For the specific information regarding the Usefulness 3 2 s Alert service, the information of the alerted 1 0 objected is retrieved from the ontology and Extendibil Accuracy given to the farm caretakers. This includes the ity method or solution to handle the alerted issue such as how to cure a disease or to prevent Sensibility spreading of a disease. This part of the service is designed to help farm caretakers to swiftly Figure 3. Average Evaluation Results on Quality handle the situation before the tamarind fruits of the Ontology in 6 Aspects are damaged. The results show that ontology experts IV. EVALUATION agreed that the ontology is in high quality, especially for both coverage and accuracy of To evaluate the proposed framework, we ontological relations for the average of 4.8 set up two experiments. The first experiment is score. It is noticeable that the experts gave the to evaluate the designed Tamarind Management lowest score as 3.8 to extendibility quality. Ontology and its accompanied production From interviewing, they mentioned that the rules. This evaluation is to ascertain the concepts in the ontology are too specific for quality of our knowledge base regarding the application and should need some editions

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 26 Phimphan Thipphayasaeng and Sant Phanichsiti for extending to other knowledge; for different type of farmland (uphill and forest instance, there are other types of pests for area) in the future to make the ontology more other , and the cultivation phases for flexible to users’ various farm setting. other plants may be different. Some of the experts also added that we should consider the

TABLE VI ASPECTS FOR EVALUATING THE DESIGNED ONTOLOGY

Ontology quality aspects to be evaluated Description Coverage regarding the scope of the work To evaluate the given classes and scope of this work To evaluate the ontology classes and relation among Correctness of contents classes regarding the domain knowledge To evaluate the ontological structure in terms of ontology Accuracy of ontological relations design and semantic representation To evaluate the terms (in Thai) chosen to represent class Sensibility of chosen terms and relation name To evaluate the overall design in case of extendibility for Extendibility of the ontology other usages including extension to other fruits and extension for handling another issue of tamarind farming To evaluate the usability of the ontology in the designed Usefulness of the ontology application

B. Application Result Evaluation  False Negative (FN): alert is not given In this section, we tried to evaluate the but there is actual invasion within 2 days or practical use of the proposed framework. The caretakers expect the invasion. focused service in this experiment was the Alert service in which took the real-time data In calculation of P, R and F1, we use (1), from deployed sensors and made decision (2), and (3), respectively. The results are then regarding ontological rules. separated into cases of disease and insect invasion, and they are given in Table VII. In this experiment, we set up the system as explained in Section 3 in three voluntary (1) tamarind farms. All farms had cultivated for tamarind fruits at least 5 years, and the farm (2) caretakers were an experienced farmer. The period of the experiment was 120 days for (3) each farm around April to July, 2019. The automatically generated alerts of disease and The ratio of all incidents of all experimented insect invasion occurred in the farms were farms was 13:40 while the alerts were counted and compared with farm caretakers’ triggered for 14:47 for disease cases and insect expectation or happening of the invasion of cases respectively. From the results, we notice disease and insect. The measurements in this that the alerts on diseases obtained the high evaluation are precision (P), recall (R) and f- score in both precision and recall. In the other measure (F1) [45] with the count criterion as hand, the alerts of insects suffered from lower follows. precision score as 0.79 in average since there were many false positive cases. However, the  True Positive (TP): alert is given and framework received a satisfying high recall in there is actual invasion within 2 days or overall average for 0.97 score. This indicates caretakers expect the invasion. that the alerts could cover most of the invasion cases in practical.  False Positive (FP): alert is given but there is no actual invasion within 2 days or In an attempt to analyze the false alerts, caretakers do not expect the invasion. especially the insect cases, we interviewed

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 27 Ontological and Sensor-Based Supporting System on Damage Prevention for Tamarind Farming: a Case Study of Phetchabun Tamarind Farms farm caretakers about it. We learned that these By asking the participants about other farms had applied on organic pesticide made services aside of the experimented Alert by natural products beforehand. This effectively service, they mentioned on lacking of the prevented the invasion of certain insects and service to help them in fertilization and soil thus made the system to false alarm on insect management. Additionally, they mentioned invasion. They also mentioned that unless they that there were other invaded insects and applied such pesticide, the alerted insects may animals in which are local pest differently occur in their farm. based on their area and yet included in the system.

TABLE VII EVALUATION RESULTS IN PRECISION, RECALL AND F-MEASURE OF THE PROPOSED FRAMEWORK

Disease Insect All Farm ID P R F1 P R F1 P R F1 A 1.00 1.00 1.00 0.76 1.00 0.87 0.80 1.00 0.89 B 1.00 1.00 1.00 0.73 0.92 0.81 0.79 0.94 0.86 C 0.86 1.00 0.92 0.87 0.93 0.90 0.86 0.95 0.90 Average 0.95 1.00 0.98 0.79 0.95 0.86 0.85 0.97 0.91

V. CONCLUSION AND FUTURE WORK In the future, we plan to extend the framework to cover farming of other plants including Smart farming has been in need to support and rice by extending the developed Thai agriculture to improve their productivity, ontology. Moreover, we will apply knowledge especially the products of locals. This work of fertilizing and soil management into the presents a method to combine the intelligence framework for coverage of cultivation process. of farm caring and the detection of real-time In addition, we plan to deploy the systems to data for supporting tamarind farming which is other tamarind farms in the area to facilitate the major products of Phetchabun, Thailand. their productivity. The use of ontology representation for collecting related information in handling REFERENCES invasion of pest and disease along with its inference rule makes the framework able to (Arranged in the order of citation in the make rational decision similar to human same fashion as the case of Footnotes.) thinking. Sensors are deployed to obtain real- time data of climate factors for precise [1] Zohrameena, S., Mujahid, M., Bagga, P., information in individual farms. By combining Khalid, M., Hasan, N., & Ahmad, N. both, intelligence and precision could be (2017). Medicinal uses & pharmacological yielded for smart farming towards tamarind activity of Tamarindus indica. World cultivation. From the evaluation, the designed Journal of Pharmaceutical Sciences, 5, ontology received an approve for coverage 121-133. regarding the scope of the work, accuracy of [2] Kuru, P. (2014). Tamarindus indica and ontological relations and correctness of its health related effects. Asian Pacific contents. In the experiment in practical use, Journal of Tropical Biomedicine, 4(9), the framework obtained an impressive recall 676-681. and f-measure score in overall average for [3] Bhadoriya, S.S., Aditya, G., Jitendra, N., 0.97 and 0.91, respectively. The result also Gopal, R., & Pal, J.A. (2011). Tamarindus shows that the framework performed the best indica: Extent of explored potential. in alerting for disease invasion. Pharmacogn Rev., 5(9), 73-81. [4] Menezes, A.P., Trevisan, S., Barbalho, S.M., & Guiguer, E.L. (2016). Tamarindus indica L. A with multiple medicinal

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 28 Phimphan Thipphayasaeng and Sant Phanichsiti

purposes. Journal of Pharmacognosy and Rajabhat University). Phytochemistry, 5(3), 50-54. [13] Snamchaiskul, C. (2015). Management [5] Uchenna, U.E., Shori, A.B., & Baba, of Sweet Tamarind Pod Borer with A.S. (2018). Tamarindus indica seeds Sufficient Economy by Participation of improve and lipid metabolism: Sweet Tamarind Community Enterprise An in vivo study. Journal of Ayurveda in Phetchabun Province (Phetchabun and Integrative Medicine, 9 (4), 258-265. Rajabhat University). [6] Sasidharan, S.R., Joseph, J.A., [14] Pearmalang, T. (2015). The Studying of Anandakumar, S., Venkatesan, V., Controlling Zeuzera coffeaeby Participation Madhavan, C.N.A., & Agarwal, A. of the Agriculturalists in Amphur Muang (2014). Ameliorative Potential of Phetchabun Province (Phetchabun Rajabhat Tamarindus indica on High Diet University). Induced Nonalcoholic Fatty Liver [15] Bakar, Z.A. & Ismail, K.N. (2013). Base Disease in Rats. Scientific World Journal, 1-10. Durian Ontology Development Using [7] Mahmudah, R., Adnyana, I.K., & Kurnia, N. Modified Methodology. M-CAIT, 206-218. (2017). Anti-Asthma Activity of Tamarind [16] Jain, S. & Tiwari, S.M. (2014). Knowledge Pulp Extract (TAMARINDUS INDICA Representation with Ontology Tools & L.). International Journal of Current Methodology. International Journal of Pharmaceutical Research, 9(102), 102-105. Computer Applications (ICACEA-2014), 1-5. [8] The Phetchabun Provincial Statistical [17] Grimm, S., Hitzler, P., & Abecker, A. Office. (2015). A situational analysis of (2007). Knowledge Representation and Sweet Tamarind production in Ontologies. Semantic Web Services: Phetchabun Province. (The Phetchabun Concepts, Technologies, and Applications, Provincial Statistical Office, Online 51-105. published Manuscript), 2. [18] Faculty of Agricutural and Industrial [9] Snamchaiskul, C. (2011). Integrated Technology Phetchabun Rajabhat University. Management of Fungi in Sweet (2014). Sweet Tamarind in Phetchabun Tamarind Orchard for Community Province. Knowledge Management of Economy Network. Farmers in Phetchabun Faculty of Agricutural and Industrial Province (Phetchabun Rajabhat University). Technology, Phetchabun Rajabhat University. [10] Srisuvoramas, B., Snamchaiskul, C., & [19] National Bureau of Agricultural Commodity Jantaramungkorn, N. (2010). Studies on and Food Standards. (2015). Pest List of the Fungal Infected Stages and the thailand. National Bureau of Agricultural Fungal Inhibitor Effect of Trichoderma Commodity and Food Standards. Retrieved sp. Fermented Medicinal Plants, from http://ippc.acfs.go.th/pest/. Azadirachtin extract and Wood Vinegar [20] Sirimongkararat, S., Saksisirat, V., Wongsorn, on Prakaithong Sweet Tamarind in D., Thongphak, D., & Ponganan, K. Vitro. Phetchabun Rajabhat Journal, 12 (2017). Diversity of Edible (2), 80-90. Insects in Community Forests (Plant [11] Chanartaepaporn, P., Srisuvoramas, B., Genetic Conservation Project) in Khon Pearmalang, T., & Snamchaiskul, C. Kaen Province. KKU Sci. J, 45(3), 551-565. (2015). Management of sweet tamarind’s [21] Tongsri, V. & Sangchote, S. (2016). pest with sufficient economy by Latent Infection of Phomopsis sp., the participation of sweet tamarind community Causal Agent of Leaf Spot on Durian enterprise of Phetchabun province (Durio zibethinus Murr.) (Phetchabun Rajabhat University). Monthong. King Mongkut’s Agricultural [12] Srisuvoramas, B. & Khoomsab, K. Journal, 34(1), 59-67. (2014). Identification and Classification [22] Matić, S., Cucu, M.A., Garibaldi, A., & of Some Fungi and the Fungal Inhibitor Gullino, M.L. (2018). Combined Effect Effect of Fermented Terminalia sp. on of CO2 and Temperature on Wheat Tamarindus indica L. (Phetchabun Powdery Mildew Development. Plant

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 29 Ontological and Sensor-Based Supporting System on Damage Prevention for Tamarind Farming: a Case Study of Phetchabun Tamarind Farms Pathol, 34(4), 316-326. Agriculture, 156, 467-474. [23] Pastirčáková, K. & Pastirčák, M. (2013). [33] Phanthuna, N. & Teerachai, L. (2017). A powdery mildew (Pseudoidium sp.) Design and Application for a Smart found on Chelidonium majus in the Farm in Thailand Based on IoT. Applied Czech Republic and Slovakia. Czech Mechanics and Materials, 866, 433-438. Mycology, 65(1), 125-132, ISSN: 1805- [34] Parisutthikul, S., Faarungsang, S., 1421. Duangjinda, M., & Thongpan, A. (2010). [24] Chariyamakarn, W., Boonbrahm, P., Web-based Information System for Supnithi, T., & Ruangrajitpakorn, T. Management of Swine Breeding Herd (2017). An Ontology-based Supporting Farm. Kasetsart Journal - Natural Science, 44. System for Integrated Farming towards a [35] Gruber, T.R. (1995). Toward principles Concept of the Sufficiency Economy. for the design of ontologies used for The KKU Science Journal, 44(4), 691-704. knowledge sharing?. International Journal of [25] Pesonen, L.A., Teye, F.K.W., Ronkainen, Human-Computer Studies, 43 (5-6), 907- A.K., Koistinen, M.O., Kaivosoja, J.J., 928. Suomi, P.F., & Linkolehto, R.O. (2014). [36] Igor, J., John, M., & Eric, Y. (1999). Cropinfra – An Internet-based service Using Ontologies for Knowledge infrastructure to support crop production Management: An Information Systems in future farms. Biosystems Engineering, Perspective. Knowledge and Information 120, 92-101. Systems, 6. [26] Agri-Map. Retrieved from http://agri- [37] Leung, N.K.Y., Lau, S.K., & Fan, J.P. map-online.moac.go.th/login. (2007). An Ontology-Based Knowledge [27] Rice - Time. Retrieved from Network to Reuse Inter-Organizational https://play.google.com/store/apps/details?id= Knowledge. xyz.ideapop.ricetime&hl=th. [38] Gulzar, Z. & Leema, A.A. (2016). An [28] LDD Soil Guide. Retrieved from ontology based approach for exploring http://www.ldd.go.th/www/lek_web/web knowledge in networking domain, .jsp?id=17837. International Conference on Inventive [29] Rice Knowledge Bank, Department of Computation Technologies (ICICT). Rice, Thailand. Retrieved from Coimbatore, 1-6. http://www.ricethailand.go.th/rkb3/index [39] Abdel-Badeeh, S. & Alfonse, M.M. .htm. (2008). Ontology versus Semantic Networks [30] Chariyamakarn, W., Boonbrahm, P., for Medical Knowledge Representation. Boonbrahm, S., & Ruangrajitpakorn, T. 12th WSEAS International Conference (2015). A Framework of Ontology based on COMPUTERS, Heraklion, Greece, Recommendation for Farmer Centered July 23-25, 2008. Rice Production. Proceedings of the [40] Ruangrajitpakorn, T., Phrombut, C., & Conference the Tenth International Supnithi, T. (2018). A Development of Conference on Knowledge, Information an Ontology-based Personalised Web and Creativity Support Systems (KICSS from Rice Knowledge Website. The 13th 2015). International Conference on Knowledge, [31] Fongngen, W. (2018). Application with Information and Creativity Support the Internet of Things Technology Systems (KICSS 2018). Control in Smart Farms Mushroom. [41] Lim, S.C.J., Liu, Y., & Chen, Y. (2015). Journal of Information Technology Ontology in Design Engineering: Status Management and Innovation, 5(1), 172-182. and Challenges. Internationnal Conference on [32] Muangprathuba, J., Boonnama, N., Engineering Design, ICED15, Politecnico Di Kajornkasirata, S., Lekbangponga, N., Milano, ITALY. Wanichsombata, A., & Nillaorb, P. [42] Kozaki, K. & Sunagawa, E. (2002). (2019). IoT and agriculture data analysis Hozo: an Ontology Development Environment for smart farm. Computers and Electronics in - Treatment of Role Concept and

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 30 Phimphan Thipphayasaeng and Sant Phanichsiti

Dependency Management. Knowledg Engineering and Knowledge Management: Ontologies and the Semantic Web Lecture Notes in Computer Science, 155-163. [43] Buranarach, M., Thein, Y.M., & Supnithi, T. (2013). A communitydriven approach to development of an ontology-based application management framework. In: Takeda, H., Qu, Y., Mizoguchi, R., & Kitamura, Y. (eds.), Semantic Technology. Springer Berlin Heidelberg, Berlin, 306- 312. [44] Likert, R. (1932). A Technique for the Measurement of Attitudes. Archives of Psychology, 140, 1-55. [45] Powers, D.M.W. (2011). Evaluation: From precision, recall and F-measure to ROC, informedness, markedness & correlation. Journal of Machine Learning Technologies, 2(1), 37-63.

International Journal of the Computer, the Internet and Management Vol.28 No.3 (September-December, 2020) pp. 20-31 31