Professional Documents
Culture Documents
Intelligent Decision Support Systems of Medicinal
Intelligent Decision Support Systems of Medicinal
Intelligent Decision Support Systems of Medicinal
Abstract— The richness of Borneo's biodiversity with the Another impact is that flooding occurs because the
potential for indigenous tribal knowledge, one of which is using barrier's forest for preventing flooding has been deforested.
various types of medicinal plants used in traditional local ethnic Frequent floods that afflict areas throughout villages and cities
medicine, especially those around forest areas. This study aims in Borneo cause skin diseases. Skin diseases (skin disorders)
to develop Borneo's medicinal forest plants decision-making are caused by fungi, germs, parasites, viruses, or infections [9]
system for skin diseases based on the International that affect anyone of all ages [10], [11]. Skin diseases can
Classification of Diseases (ICD-10 version 2016). The attack all or certain parts of the body and can worsen the
determination of medicinal plants for the treatment of skin
patient's health condition if not treated seriously [12]. Skin
diseases is based on criteria types of plant species, how to
disorders often occur due to factors such as climate,
process, how to use, and plant parts used. This research resulted
environment, place of residence, unhealthy living habits,
in a decision support system intelligence software product for
determining of 94 dataset medicinal forest plants for skin allergies [13] and others.
diseases using the AHP method for weight determination The potential for medicinal plants in the Borneo forest area
(priority), and WASPAS for preference. The implementation of is very diverse [14], both those that have been exploited by the
the AHP-WASPAS method in case studies of medicinal forest people around the area and those that have not been utilized.
plants decision shows that the user's subjectivity in weighting
Various studies have been conducted to explain that the
and decision-making criteria affects the recommended
Borneo forest area has the potential for various medicinal
preference values. Furthermore, the multi-criteria analysis
plants. The currently recorded potential has not shown the
method approach to decision making in applied case studies is
less objective because the knowledge base of alternatives and
potential of Borneo medicinal plants as a whole, but it can
criteria is complex. describe the potential of medicinal plants in certain forest
areas only [14]. The potential for medicinal plants is scattered
Keywords— forest, medicinal-plants, skin-disease, AHP- in various forest areas including conservation areas such as
WASPAS, Borneo research forest areas, national parks, protected forests and
other forest areas. Types of medicinal plants that have been
I. In t r o d u c t io n identified and documented through various studies [14], [15],
The utilization of various forest plant types for the one of which is the types of forest plants for the treatment of
treatment of disease is one of the local cultural heritages that skin diseases.
are preserved by the indigenous tribe of Borneo [1], [2]. The study purposes to develop a decision support system
Treatment of minor and severe diseases uses ingredients for determining Medicinal Forest Plants for skin diseases
derived from certain types of plants found around gardens and based on the International Statistical Classification of
in forests. Currently, the use of medicinal plants or jamu is an Diseases [16] (ICD-10 version 2016). The main source of data
alternative for people to maintain health and treat disease [3], collection on medicinal plants is obtained from the Research and
this is because the use of medicinal plants or herbs is not only Development Centre for Natural Resources Technology, East
affordable but also does not cause side effects when using Borneo Indonesia. Determining the weighting importance of
modern medicinal plants derived from chemicals [4]. medicinal plants using the Analytical Hierarchy Process
Most of the medicinal plants are used by indigenous (AHP) and the Weighted-aggregated Sum Product-assessment
people in rural areas, especially in areas where public health (WASPAS) for preference decision recommendations.
facilities are not yet reached. For their daily needs, they often Research contribution is intended as an added value to science
take plants as medicinal raw materials directly from natural in the pharmaceutical sector, particularly in the use of
forests [5]. However, in the last few decades, there has been medicinal forest plants for the treatment of skin diseases based
forest destruction and/or deforestation of forests in Borneo [6], on information technology and decision support systems.
which has changed the function of forests into activities and Provide recommendations for the selection of drugs for the
human interests [7]. Without realizing it, deforestation or right skin diseases related to the processing method, method
forest destruction have impacts such as climate change [8]. of use, and parts of medicinal plants used for treatment,
Authorized licensed use limited to: Raytheon Technologies. Downloaded on May 25,2021 at 04:27:07 UTC from IEEE Xplore. Restrictions apply.
making it easier for the community to avoid misconceptions 1) Analytic Hierarchy Process (AHP), is method from
properties of the forest medicinal plants. Thomas L. Saaty and supports decision-makers to find the
best alternative from the many elements of choice [25]. The
II. M a t e r ia l s
AHP initial stage in determining element priority, comparing
A. Skin Disease elements in pairs to represent the relative importance of one
Skin disease in this study is defined as a skin disorder element to another [26] through a pairwise comparison
caused by fungi, germs, parasites, viruses or infections that matrix.
can affect anyone of all ages [10]. Disorders of the skin often 2) Weighted Aggregated Sum Product Assessment
occur due to factors such as climate, environment [17], (WASPAS), is a combination of weighted sum model and
habitation, lifestyle, allergies and other s [18]. Skin disease is weighted product model [27], [28]. The WASPAS method is
increasingly developing, this is evidenced by Indonesia's expected to provide better results to determine the DSS of
health profile in 2015 which shows that skin and subcutaneous
forest medicinal plants with properties for skin diseases.
tissue disease ranks third of the 10 most common diseases of
outpatients at hospitals in Indonesia based on the number of III. M e t h o d o l o g y
visits, namely 192,414 visits, new case visits 122,076 visits
while the old case was 70,338 visits [19], [20]. The A. Data Collection Methods
classification code for skin diseases according to ICD-10 Several data collection methods were applied, among
version:2019 [21] and Taxonomy of International Code of others, through literature studies and various data sources for
Botanical Nomenclature (ICBN) [22]. forest medicinal plants, and other. The main data source on
medicinal plants is obtained from the results of research
B. Medicinal Forest Plants publications and books related to forest medicinal plants. In
Medicinal plants in Kalimantan forests are not only woody addition, the method of interviewing expert sources from
plants but also non-timber plants with various habitus, namely several natural resource conservation organizations, national
trees, shrubs, herbs, lianas and ferns [23]. Medicinal plants in parks, botanical gardens in East Kalimantan Province.
forest areas in Kalimantan have the most tree and shrub B. Decision Making: Elements and Variables
habitus [24]. The recorded of potential medicinal plant in
some forest areas in Kalimantan [1] according to Noorhidayah Forest medicinal plants are an alternative treatment for
et al., are shown in "Table I". skin diseases analyzed based on 4 criteria, ie forest plant
Some studies have even examined the ethnobotany of species types(C1): Tree, Shrubs, Liana, Bush, Herbs, how to
certain tribes in Kalimantan, including the use of medicinal process(C2): Pounded, Boiled, Direct, Shredded, Burned, how
plants for treatment [14]. There is a tendency that the types of to use(C3): Eaten, Stuck, Drunk, Smeared, Bathe and parts of
medicinal plants found in forest areas in Kalimantan are the forest plants(C4): Leaf, Root, Fruit, Rhizome, Stem, Sap,
types used by the tribes who inhabit the area [14]. Flower, Branches, Seed.
C. Intelligent Decision Support System (DSS) Analysis C. Decision Analysis Methods: AHP-WASPAS
The knowledge-based system management model of The design of the decision-making process for
medicinal forest plants is packaged in the framework of an determining medicinal forest plants for the treatment of skin
intelligent DSS that recommend several alternative decisions diseases combines the AHP-WASPAS method. AHP is used
to become the preference for forest plant species desired by to determine element weights(criteria) while WASPAS is
the user, based on several criteria such as how to process, how used to determine preference values [29]. An overview of the
to use, the part of the plant, and plant types. The analytical flowchart decision-making process for determining medicinal
method the decision making of the forest medicinal plants for forest plants for the treatment of skin diseases is shown in
skin disease are: "Fig. 1"
TABLE I. t he P o t e n t ia l M e d ic in a l Pl a n t s in s o m e Fo r e s t
AREAS IN KALIMANTAN[ 1]
355
Authorized licensed use limited to: Raytheon Technologies. Downloaded on May 25,2021 at 04:27:07 UTC from IEEE Xplore. Restrictions apply.
1) Elements Priority and Pairwise Comparisons Matrix: A. Results:Collection Data
The elements priority stage is making pairwise comparisons The main source of data on medicinal plants is obtained
that is comparing elements in pairs according to the available from the results of research publications and books related to
criteria [30]. Pairwise comparison matrix is filled using medicinal forest plants. In addition, the method of
numbers to represent the relative importance of an element interviewing expert sources from several natural resource
with other elements[30]. Using the Saaty’s scale[31], and conservation organizations, national parks, botanical gardens,
Vargas [25] the equation use "(1)". in East Kalimantan Province. Data collection results in “Fig.
2” for the knowledge base on skin disease treatment obtained
94 medicinal forest plants consisting of 28 tree species, Herb
1 ri2 ... r±n of 16 species, Liana of 8 species of, shrubs of 31 species, and
1/ r i2 1 ••• r2n Bush of 11 species. Data for medicinal forest plants collected
( 1) for the 4 types branches, 8 types roots, 56 leaves, 6 rhizomes,
. 1 6 fruits, 7 saps, 2 flower and 3 stems. Furthermore, a
1/ rin
screenshot of the User interface on the decision support
system is shown in "Fig 3".
Where, i, m = 1, 2 .... n = index criteria
Determining Relative Priority, calculate its value in
equation use “(2”).
Y? r.
P rio rity — i = - - (2)
n
CI = . max n (4)
n—1
Fig. 3. DSS GUI of the Medicinal forest plants for skin diseases
356
Authorized licensed use limited to: Raytheon Technologies. Downloaded on May 25,2021 at 04:27:07 UTC from IEEE Xplore. Restrictions apply.
Fig. 3. shows the medicinal forest plant data recorded in TABLE II. M e d ic in a l Fo r est Pl ant s for Sc a b ie s D is e a s e
the database system. The attribute of medicinal forest plant Plant
Plant Part of How to
data contains information on the Latin and local name of the Local Latin Name
types plants process
plant, taxonomy (family), how to process is, how to use, the Name
plant parts used are efficacious for the treatment of skin Camellia
Tea tree oil Tree Sap Direct
sinensis
diseases, and others.
Aloe vera Aloe Vera Herbs Sap Direct
B. Results: DSS Medicinal Forest Plant fo r Skin Disease Cajanus Cajan
Pigeon pea Tree Leaf Pounded
L. Millsp.
The implementation of the decision analysis method in the Syzygium
intelligent system for determining medicinal plants for skin Bay-leaf Tree Leaf Pounded
Polyanthum
diseases is designed in a graphical user interface (GUI) in the Mudar Calotropis
Tree Leaf Pounded
disease selection process shown in “Fig. 4”. Interface system plant Gigantea (L.) D
in "Fig. 4" shows the menu options for the type of disease user Neem, Azadirachta
Tree Leaf Pounded
selected, there are 4 lists ie. How to use (drunk, eaten, Nimtree indica A. Juss
smeared, stuck and bathed), how to process (direct, burned, Curcuma
Turmeric Shrubs Rhizome Pounded
pounded, boiled, shredded), and a list of plant parts used to domestica Val.
treatment (roots, stems, branches, leaf, sap, rhizome, fruits,
flowers and seeds). Besides that, you will also list the types of
plants such as trees, shrubs, lianas, bushs and herbs. An
example is the use of the DSS system to determine Forest
plants in the treatment of "scabies" if the user selects a
checklist as shown in "Fig. 4".
In "Fig 4" if the user selects scabies, the system provides a
checklist of options for each criterion, for example, the user
chooses how to use medicinal plants "smeared", chooses the
Fig. 6. The Max Eigenvalues o f the normalization matrix
checklist "direct, and pounded" how to process, and selects all
parts and types of plants that treat the Scabies disease. Thus,
In "Table II" obtains there are 7 medicinal forest plants for
the DSS system analyzes using the AHP-WASPAS method
the treatment of scabies, 5 types of trees, 1 type of shrub, and
and recommends Medicinal Forest Plants that can treat
1 herb. The plant parts used are 2 Sap, 4 Leaves, and 1
Scabies disease according to the criteria selected by the user.
Rhizome. As for the processing, all the plants are Pounded.
The level of pairwise comparison weight recommendation
Each criterion and forest plant element is weighted as the best
from experts from the Center for Research and Development
alternative recommendation from the DSS system. The next
of Conservation Technology for Natural Resources of East
stage is determining the relative priority value for each
Kalimantan Province which determines the level of
criterion based on "Fig. 5" use "(2)" so that the value is shown
importance of criterion (C) to other criteria in shown in “Fig.
in "Table VI". Determination of the Max Eigenvalues of the
5”. In case the user selects "Scabies", then the system
pairwise comparison matrix in "Fig. 6" use "(3)". The
processes forest plants that can treat. In the DSS system, there
calculation results obtained the value of k_max = 4.088726.
are 7 knowledge bases of 94 forest plants for the treatment of
Whereas for the calculation of the consistency ratio use "(4)",
scabies are shown in "Table II".
for the random index value in "Fig. 6" the CR value is obtained
= 0.03286.
Hierarchy consistency check for CR value <0.1 (0.03286
<0.1) so that this calculation is stated to be consistent, then the
relative priority value for each criterion can be used as the
weight of the criteria. The WASPAS method to determine the
priority of alternative medicinal forest plants for skin diseases,
the determination of the weight based on the ratio value of the
alternatives number, the results of weighting determining are
shown in the "Table III".
The weighting in the "Table IV" is used to convert the data
set into an alternative decision matrix table in determining
alternative priorities.
Fig. 4. GUI DSS o f medicinal forest plant for skin disease TABLE III. W e ig h t De c is io n M a t r ix f o r Sc a b ie s
357
Authorized licensed use limited to: Raytheon Technologies. Downloaded on May 25,2021 at 04:27:07 UTC from IEEE Xplore. Restrictions apply.
TABLE IV. W e ig h t s : Pl a n t t y p e , Pa r t U s e d , h o w t o u se,a nd This research resulted in a decision support system
H o w t o Pr o c e s s
intelligence software product for determining medicinal forest
Sub-Criterion
Forest
Freq. Weights
Eigen’s plants for skin diseases using the AHP method for weight
Plants Priority determination (priority), and WASPAS for preference. The
Tree 5 0.714286 implementation of the AHP-WASPAS method in case studies
Plant Types 0.1898
Shrub 1 0.142857 of medicinal forest plants decision shows that the user's
Herb 1 0.142857
subjectivity in weighting and decision-making criteria affects
Leaf 4 0.571429
Plant Part 0.0887 the recommended preference values. The multi-criteria
Rhizome 1 0.142857
Sap 2 0.285714 analysis method approach in the case study of medicinal forest
Direct 2 0.285714 plant data exploration for decision making in the analysis is
How to process 0.2346
Pounded 5 0.714286 less objective and requires a knowledge-based analysis
How to Use Smeared 7 1 0.4867 method where alternatives and criteria are complex. Future
research is applied to the expert system analysis method
approach which is expected to optimize the inference engine
for more objective results.
Ac k n o w l ed g m en t
358
Authorized licensed use limited to: Raytheon Technologies. Downloaded on May 25,2021 at 04:27:07 UTC from IEEE Xplore. Restrictions apply.
[14] J. A. Widians, M. Wati, and A. Tejawati, “Biodiversity Information [24] E. Budiman, M. Jamil, U. Hairah, H. Jati, and Rosmasari, “Eloquent
System for Management of Medicinal Plants Data Tropical Rainforest obj ect relational mapping models for biodiversity information system,”
Borneo,” Int. J. Eng. Technol., vol. 7, no. 4.44, pp. 31-36, 2018, doi: in 2017 4th International Conference on Computer Applications and
10.14419/ijet.v7i4.44.26858. Information Processing Technology (CAIPT), Aug. 2017, pp. 1-5, doi:
[15] Y. W. dan Z. Azham, “Inventarisasi Jenis Tumbuhan Yang Berkhasiat 10.1109/CAIPT.2017.8320662.
Sebagai Obat Pada Plot Konservasi Tumbuhan Obat Di KHDTK [25] T. Saaty and L. Vargas, Models, methods, concepts & applications o f
Samboja Kecamatan Samboja Kabupaten Kutai Kartanegara,” the analytic hierarchy process. 2012.
AGRIFOR, 2017. [26] L. Della Spina, “Adaptive sustainable reuse for cultural heritage: A
[16] NCBI, “ICD-10 version 2016,” National Center fo r Biotechnology multiple criteria decision aiding approach supporting urban
Information. https://www.ncbi.nlm.nih.gov. development processes,” Sustain., 2020, doi: 10.3390/su12041363.
[17] D. M. Khairina, H. R. Hatta, R. Rustam, and S. Maharani, “Automation [27] E. Ilbahar, S. Cebi, and C. Kahraman, “Assessment o f renewable
Diagnosis of Skin Disease in Humans using Dempster-Shafer Method,” energy alternatives with pythagorean fuzzy WASPAS method: A case
2018, doi: 10.1051/e3sconf/20183111006. study of Turkey,” 2020, doi: 10.1007/978-3-030-23756-1_106.
[18] W. H. Organization and Others, “Epidemiology and management of [28] S. Chakraborty, E. K. Zavadskas, and J. Antucheviciene, “Applications
common skin diseases in children in developing countries,” Geneva: o f WASPAS method as a multi-criteria decision-making tool,” Econ.
World Health Organization. 2005. Comput. Econ. Cybern. Stud. Res., 2015.
[19] Profil Kesehatan Indonesia, ProfilKesehatan R I 2015. 2016. [29] E. Budiman, N. Dengen, Haviluddin, and W. Indrawan, “Integrated
[20] N. Dengen, Haviluddin, L. Andriyani, M. Wati, and F. Alameka, multi criteria decision making for a destitute problem,” in 2017 3rd
“Medicine Stock Forecasting Using Least Square Method,” in 2018 International Conference on Science in Information Technology
2nd East Indonesia Conference on Computer and Information (ICSITech), 2017, p. 342-347, doi: 10.1109/ICSITech.2017.8257136.
Technology (EIConCIT), 2018, pp. 100-103, doi: 10.1109/- [30] M. Irfan Ramadhan, M. Zarlis, and B. B. Nasution, “Performance
EIConCIT.2018.8878563. analysis o f combination o f fuzzy analytic hierarchy process (FAHP)
[21] WHO, “ICD-10 Version:2019 Chapter XII Diseases o f the skin and algorithms with preference ranking organization method for
subcutaneous tissue,” https://icd.who.int/browse10/2019/en. enrichment evaluation algorithm (PROMETHEE II) in the ranking
process to determine the increase in employee class,” 2020, doi:
[22] E. Budiman and S. N. Alam, “Database: Taxonomy o f plants
10.1088/1757-899X/725/1/012107.
Nomenclature for borneo biodiversity information system,” in 2017
Second International Conference on Informatics and Computing [31] K. Chen, X. Huang, Y. Zhang, and Q. Ai, “Research on rehabilitation
(ICIC), 2017, pp. 1-6, doi: 10.1109/IAC.2017.8280642. assessment methods based on human gait and sEMG,” Cogent Eng.,
vol. 3, no. 1, 2016, doi: 10.1080/23311916.2016.1220113.
[23] U. Hairah, A. Tejawati, and F. Agus, “Borneo biodiversity: Exploring
endemic tree species and wood characteristics,” in 2017 3rd [32] E. K. Zavadskas, Z. Turskis, J. Antucheviciene, and A. Zakarevicius,
International Conference on Science in Information Technology “Optimization o f weighted aggregated sum product assessment,”
(ICSITech), 2017, p. 435-440, doi: 10.1109/ICSITech.2017.8257152. Elektron. ir Elektrotechnika, 2012, doi: 10.5755/j01.eee.122.6.1810.
359
Authorized licensed use limited to: Raytheon Technologies. Downloaded on May 25,2021 at 04:27:07 UTC from IEEE Xplore. Restrictions apply.