Implementation Weighted Product Method for the Best Carrot Seed

Recommendations

Jantianus

, E. E. Surbakti

, R. W. Sembiring

, P. Silaen

and Khairul

Medan State Polytechnic, Jln. Almamater No. 1, Medan, North Sumatera, Indonesia

Multimedia Nusantara University, Jl. Scientia Boulevard, Tangerang, Banten, Indonesia

Bunda Mulia University, Jl. Ancol Barat IV, Tangerang, North Jakarta, Indonesia

khairulhasyar@polmed.ac.id

Keywords: Artificial Intelligence, Carrot Seeds, Decision Support System, Mount Sinabung Refugee Farmers, Weighted

Product Method.

Abstract: So far, in determining and looking for criteria for good/excellent seeds, farmers do not have the right tools or

methods. The selection of good/superior seeds is done based on personal experience and views, even some of

the Sinabung refugees who cultivate carrots do not know the source of the seeds they use. Varieties of carrots

are also very many kinds, almost 40 kinds of varieties. Each variety has different growing criteria and

requirements. Thus, this becomes an additional difficulty factor in determining the quality of the type of carrot

seed to be planted. This of course affects the production of carrots at harvest. Harvest production is not

proportional to the area of land used. Based on the problems above, the search for superior seeds using a

decision support algorithm can provide recommendations. This algorithm will provide recommendations for

the quality of carrot seeds based on the criteria or characteristics that must be prepared in the selection of

plants. The algorithm developed uses the Weighted Product (WP) algorithm. The WP method is a

recommendation method with weighting against predetermined criteria. The weights given are based on

research results and expert experience. From the results of this trial, several recommendations for prospective

carrot plants are sorted based on the results of the algorithm calculations and can be used as superior seeds

that produce greater production. The results of the study are expected to because of this research, the system

has succeeded in providing recommendations for the best carrot seeds that can help increase carrot production

and help economic resilience for farmers displaced by Mount Sinabung.

1 INTRODUCTION

Tanah Karo, which is well-known as a vegetable

producer, is the main source of vegetable suppliers in

North Sumatra, even vegetables are exported to the

national region and abroad. Nowadays, many farmers

switch to carrot plants for various reasons, such as the

planting age of only three months, practical ways of

working and the wide export market in various cities

in Indonesia, such as Jakarta, Bandung, Surabaya,

Bali and even Papua. Based on previous research,

farmers can produce Rp. 7,450,50/ha (Sundari, 2011).

The age of carrots until the harvest period is 2.5 - 3

months, judging from the results that can be obtained

by farmers in growing carrots, of course this is very

supportive of their economic resilience.

Most of the farmers in the land are still farming

semi- modern, using tractor tillage, spraying machine

showers and mostly with human labor. No one has yet

fully utilized modern farming methods such as

utilizing an IT system. This the author found in the

relocation of karo in the relocation of refugees from

Mount Sinabung in the village of Nangbelawan.

Things that need to be considered in carrot cultivation

to get maximum production, is the optimum

temperature for carrot plant growth is 15-21oC. This

temperature is suitable for the growth of roots and the

top of the plant so that the color and shape of the roots

can be optimal (Simon, 2019). Soil that is suitable for

growing carrots is soil that is well drained, rich in

organic matter and fertile with an altitude of 1200-

1500 m above sea level. Sandy loam soil is suitable

for carrot cultivation because it is easy for root.

Can reach optimal length and size. This plant can

grow well in soil with a pH of 5-8 (Putra, 2018). Soil

moisture is very important for the growth of carrot

814

Jantianus, ., E. Surbakti, E., W. Sembiring, R., Silaen, P. and Khairul, .

Implementation Weighted Product Method for the Best Carrot Seed Recommendations.

DOI: 10.5220/0010954100003260

In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 814-820

ISBN: 978-989-758-615-6; ISSN: 2975-8246

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

plants, including during seeding to obtain seeds with

uniform growth and fast growth after planting in the

field. Apart from the above requirements, good seeds

cannot be separated from the success of obtaining

maximum production (Singh, 2018). The fact that the

authors found at the location of this research,

generally the farmers follow the correct procedures in

farming, especially in the selection of good seeds.

Determining and searching for criteria for

good/excellent seeds, the farmers do not have the

tools or methods to get them. The selection of

good/superior seeds is done based on personal

experience and views, even some of the Sinabung

refugees who cultivate carrots do not know the source

of the seeds they use (Putra, 2018); (Singh, 2018).

This of course affects the production of carrots at

harvest. Harvest production is not proportional to the

area of land used. Based on the above problems, the

search for superior seeds uses decision support

algorithms and recommendations. This algorithm will

provide recommendations for the quality of superior

carrot seeds based on the criteria or characteristics

that must be prepared in the selection of plants. The

algorithm developed uses the Weighted Product (WP)

algorithm.

The WP method is a recommendation method

with weighting against predetermined criteria

(Linden, 2017). Previous research conducted by for

the selection of instant cameras, in this study

concluded that the final result of the application

evaluation was the overall value of system user

satisfaction which obtained a percentage score of

87.98% with a very good predicate (Feng, 2019). The

second research is a study for the selection of

ornamental plants, in this study it was concluded that

with a system success percentage of 84,409% the

system was considered successful in implementing

the WP method and the criteria used in the

recommendation system were appropriate. The WP

method is used in the design of this recommendation

system because it is considered to have fast

computing and is suitable for making ornamental

plant recommendations. The advantages of this WP

method provide clarity on the weights of costs and

benefits on each criterion (Linden, 2017). This study

determines the variables or criteria to determine the

characteristics of superior carrot seeds, then the

process of determining superior seeds is carried out

using the WP method. From the results of trials

conducted after the harvest period, the production of

superior carrot seeds will be compared with seeds

without the WP process. From the results of this trial,

it can be seen whether superior seeds can produce

greater production or not (Singh, 2018).

In general, carrot farmers displaced by Mount

Sinabung do not have standards in procuring superior

seeds in their agricultural business. Most of them get

carrot seeds from unknown sources. Usually buy

seeds from traders. Some of them also produce their

own seeds but do not meet proper standards in

producing superior seeds. As a result, crop yields are

not optimal, the area of land planted is not

proportional to harvest production. The crop yields

are not optimal, the impact of Covid-19 has made the

price of carrots low, due toa lack of demand both

locally and from Jakarta and Java. As a result, it

greatly affects the economic resilience of the Mount

Sinabung refugees. One of the appropriate solutions

to overcome the above problems is to produce

superior seeds from carrots themselves using the WP

method, for that it can be stated that the problem

formulation of this research is How to Design a

System Implementation to Get Superior Carrot Seeds

with a Weighted Product Algorithm?

2 LITERATURE REVIEW

2.1 Carrot Seeds

Basically, the carrot varieties commonly consumed

by the world's population are of many types, both in

shape and color, not just one type as we usually find

in Indonesia (Simon, 2019). To get maximum

production in planting carrots, it is necessary to know

the stages that must be done in planting carrots. First,

efforts should be made to use superior seeds. From a

land that has been planted with carrots and before the

post-harvest, a search for carrots can be carried out

which will be used as superior seeds. The steps that

can be taken in finding sources of superior seeds are

as follows (Marpaung, 2017):

Age after planting day at least 100 days

The tuber texture is straight, dense

The thickness of the tuber diameter>=3 cm

Glowing bright reddish

The shape of the carrot leaves is straight and

bright

While the method of seeding carrots, as follows:

Carrot leaves cut to about 10 cm

Roots that have been selected, cut in thirds

The land is loosened, sown with compost

Made a bed for planting seeds

Plant tubers with about 50 cm

Roots that will be used as seeds must be selected

properly; the planting period is at least 100 days.

Roots that are not old enough will easily rot and get

Implementation Weighted Product Method for the Best Carrot Seed Recommendations

815

disease. To know the planting period of carrots can be

known from the date of planting. In addition, from the

physical texture, whether it has hardened and looks

ripe, it can be a reference that the carrot is old enough.

Physical texture that is straight, dense, and shiny can

be seen after the carrots that are old enough are

removed.

2.2 Weighted Product Method

The research model uses the Weighted Product (WP)

method in determining a decision based on several

attributes. This method requires the decision maker to

determine the weight for each attribute. WP evaluates

m alternatives Ai (i=1,2,..,m) against a set of

attributes Cj (j=1,2…,n) where each attribute is

independent of one another (Linden, 2017). In the WP

method normalization is still carried out, where the

rating of each attribute must be raised to the first

power with the weight of the attribute in question.

3 SYSTEM DESIGN AND

METHODOLOGY

3.1 Research Stages

The following is the order from the beginning to the

conclusion of the research to be carried out.

Figure 1: Research Stages.

In Figure 1. Research Stages, the first step is to

collect initial data. This research is a case study, so

look at the phenomena that occur in the object or

place

of research. Next, formulate existing problems

observe reality with a literature review or actual

science

and theory. The result of the literature

review is to

determine the criteria for superior

seeds from carrot

plants and computer computational

algorithms that can

be used to solve problems.

The next process is to find the value of the criteria

for prospective seeds for computational

calculations

using the WP algorithm to produce

recommendations

for the quality of prospective seeds.

These prospective

seeds will be planted and

observed to measure the

results of the

recommendations. The final stage is to

reason about

the results of plants that have grown to

draw

conclusions.

3.2 System Design

Figure 2: User Dashboard Mockup.

Figure 2. User Dashboard Mockup is the initial user

interface design when opening the website. On this

page there is a button that functions to start the

recommendation process, after the button is pressed,

a stepper will appear to guide the user in filling in the

desired criteria weights. After all processes are

complete, the recommendation results will appear to

replace the stepper component.

Figure 3: Admin Dashboard Mockup.

Figure 3. Admin Dashboard Mockup is an interface

design for the results page on the admin dashboard.

This page will display the 5 criteria for the best carrot

seed varieties, pictures of carrot plants and the final

WP weight value of the existing varites. At the top

there is a button to display the modal that contains the

recommendation calculation process in detail.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

816

3.3 Criteria and Weighted Scale

The following are the criteria and the value of each

criterion weight designed in this study:

Table 1: Criteria and Wighted Scale.

ode Criteria Weighted Description Weighted

Score

C1 Root

Texture

1-5 Benefit 1-5

C2 Root

thickne

1-5 Cost 3,8,11,15,18

C3 Root

Color

1-5 Benefit 1-5

C4 Leaf

Colo

1-5 Benefit 1-5

C5 Plant

Age

1-5 Cost 59,69,79,90,1

Based on the results of interviews with

farmers/experts who have done planting and research

for more than 20 years, there are 5 criteria in

determining the quality of candidate carrot seeds. C2

and C5 are cost criteria because the root thickness and

plant age, the higher the cost for planting and land

use. so that the resulting value is higher, but the

weight of the criteria is getting lower in the WP

calculation (Ameliana, 2019).

4 RESULT AND DISCUSSION

4.1 Implementation of WP

In choosing the best carrot seed candidate to be the

best carrot seed recommendation based on 5 criteria

inputted by the user, 10 alternatives were used. One

of these alternatives will be selected and then sorted

based on the final weighted score of the WP. Each

alternative is given the necessary criteria and weights

in performing calculations so that the results obtained

are as follows:

Table 2: Description of Criteria.

Criteria Description

Root

Texture

The texture of the tuber plays an

important role in the growth

resistance and strength of the carrot

plant against pests or temperatures.

Root

thickness

The thickness of the tubers is

proportional to the texture of the

tubers to provide strong resistance to

the would-be carrot plant.

Root

Color

The color of the tubers affects the

criteria for carrot plants. A good tuber

color is a bright color and reddish or

yellow depending on the carrot

variety produced.

Leaf

Color

Leaf color is an important factor. Leaf

color affects the quality ofthe carrot

tubers produced.

The selected alternatives from various varieties of

carrot plants have the following criteria weights:

Table 3: Weighted Scale of Each Alternatives.

Code Alternatives C1 C2 C3 C4 C5

A1 Imperator 5 11 5 5 110

A2 Chantenay 4 8 4 4 79

A3 Danvers 5 5 5 3 89

A4 Mini Carrot 5 5 3 2 59

A5 Nantes 4 11 5 5 110

A6 Hercules 5 15 3 5 69

A7 Oxheart 3 15 1 5 110

A8 Red-cored 2 15 3 2 79

A9 Merida 4 8 1 1 110

A10 Rainbow 3 11 5 4 110

Next, the process of calculating the WP based on the

weight of the criteria inputted by the user. From the

results of the weights of the inputted criteria,

normalization of the weights of each criterion will be

sought.

Implementation Weighted Product Method for the Best Carrot Seed Recommendations

817

Table 4: Input Level Criteria By User.

ode Lv.1 Lv.2 Lv.3 Lv.4 Lv.5

Branch

two

Double

bend

One

Bend

Not

straight

Straight

unbranc

C2 1-3 cm 4-5

6-8

9-11

12-15

Pale

Yellow

Less

bright

Pale

red

bright

and

less re

bright

and red

Dark

green

brown

Dark

Green

and not

fresh

Not

Fresh

and

Dark

Green

Fresh

and

Dark

Green

Fresh

and

Light

Green

C5 50-59

days

60-69

days

70-79

days

80-89

days

90-110

days

In this calculation simulation, the level value

inputted by the user is:

C1 : Level 5, straight unbranched

C2 : Level 4 , 9-11 cm

C3 : Level 5, bright and red

C4: Level 5, fresh and light green

C5: Level 5, 90-110 days

Then the calculation starts from weight

normalization, the first step is to normalize the

weights of the criteria that have been entered. The

normalization process can be seen in table:

Table 5: Normalization Process.

Code Normalization Result

C1 5/(5+2+5+5+1) 0.278

C2 2//(5+2+5+5+1) 0.111

C3 5/(5+2+5+5+1) 0.278

C4 5//(5+2+5+5+1) 0.278

C5 1/(5+2+5+5+1) 0.056

Next the normalization process is complete, then

the S vector value is calculated. The calculation

process is carried out by raising the alternative weight

value to the normalized weight value, for the weight

with the cost rank attribute to be negative while the

benefit attribute to the positive rank. The process of

calculating the value of the vector S can be seen in

Table 6. Vector S Calculation Process.

Table 6: Vector S Calculation Process.

No. Calculation Vector S Result

(50.278) (11-0.111) (50.278) (50.278)

(110-0.056)

2.256

(40.278) (8-0.111) (50.278) (40.278)

(79-0.056)

1.976

(50.278) (5-0.111) (50.278) (30.278)

(89-0.056)

2.162

(50.278) (5-0.111) (30.278) (20.278)

(59-0.056)

1.715

(40.278) (11-0.111) (50.278) (50.278)

(110-0.056)

2.12

(50.278) (15-0.111) (30.278) (50.278)

(69-0.056)

1.94

(30.278) (15-0.111) (10.278) (50.278)

(110-0.056)

1.209

(20.278) (15-0.111) (30.278) (20.278)

(79-0.056)

1.157

(40.278) (8-0.111) (10.278) (10.278)

(110-0.056)

0.898

S10

(30.278) (11-0.111) (50.278) (40.278)

(110-0.056)

1.839

After getting the results from the calculation of vector

S, the next process is calculating the value of vector

V by dividing each vector S by the total sum of all

vectors S. The calculation of vector V can be seen in

Table 7. Vector V Calculation process:

Table 7: Vector V Calculation Process.

No. Calculation Vector V

Resul

2.256/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.13

1.976/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.114

2.162/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.125

1.715/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.099

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

818

2.12/(2.256+1.976+2.162+1.715+2.1

2+1.94+1.209+1.157+0.898+1.839)

0.122

1.94/(2.256+1.976+2.162+1.715+2.1

2+1.94+1.209+1.157+0.898+1.839)

0.112

1.209/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.07

1.157/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.067

0.898/(2.256+1.976+2.162+1.715+2.

12+1.94 +1.209+1.157+0.898+1.839)

0,073

V10

1.839/(2.256+1.976+2.162+1.715+2.

12+1.94+1.209+1.157+0.898+1.839)

0.106

Based on the results of calculations and results of

sorting out the WP method above that carrot seeds

which are selected to be the best carrot seeds is:

Table 8: Weighted Scale of Each Alternatives.

Code Alternatives Result Rank

A1 Imperator 0.13 1

A2 Chantenay 0.114 4

A3 Danvers 0.125 2

A4 Mini Carrot 0.099 7

A5 Nantes 0.122 3

A6 Hercules 0.112 5

A7 Oxheart 0.07 9

A8 Red-cored 0.067 10

A9 Merida 0,073 8

A10 Rainbow 0.106 6

4.2 Implementation the Website

Figure 4: User Dashboard Website.

The website page for the carrot seed

recommendation system is divided into two,

namely for users and admins. The user only has a

feature to search for the best weight of carrot seeds,

while the features that the admin has are adding

alternative data and seeing detailed calculations

from searching for the best weight of carrot seeds.

Figure 5: Admin Dashboard Website.

5 CONCLUSION

Based on the results of research that has been

successfully carried out, the conclusions of this study

are as follows:

Implementation weighted product method for the

best carrot seed recommendations is determined

based on 5 criteria’s: root texture, root thickness,

root color, leaf color and plant age.

The process of selecting carrot plants uses the

Weighted Product (WP) method which helps in

making decisions from several alternatives that

must be taken by considering the criteria.

This decision support system was developed in

the

form of a website so that users can easily see

the

results of wp calculations and see pictures of

carrot

seed recommendations.

For further system development, giving weights

alternatives and criteria can use fuzzy values.

Implementation Weighted Product Method for the Best Carrot Seed Recommendations

819

addition, the weight of the criteria can also be

obtained using the Analytical Hierarchy Process

(AHP) questionnaire, not just interviews so that

can consist of several sources people

(Surbakti,

2019).

REFERENCES

Marpaung, A. E., Karo, B., & Tarigan, R. (2017).

Peningkatan Produksi dan Mutu Benih Wortel (Daucus

carota) Varietas Lokal Melalui Pemangkasan Cabang

dan Pemupukan Boron (Increasing the Production and

Quality of Carrot Seed Local Variety Through Branch

Pruning and Boron Fertilization). Jurnal Hortikultura,

27(1), 45-54.

Kasim, A. A., & Harjoko, A. (2014, June). Klasifikasi citra

batik menggunakan jaringan syaraf tiruan berdasarkan

gray level co-occurrence matrices (GLCM). In Seminar

Nasional Aplikasi Teknologi Informasi (SNATI) (Vol.

1, No. 1).

Ameliana, W. (2019). Implementation of weighted product

method in the decision support system of university

selection in Australia. In Proc. Int. Conf. IT, Commun.

Technol. Better Life, ICT4BL (pp. 61-70).

Singh, B. K., Sharma, N., Dubey, S. K., Sharma, J. P.,

Sharma, A., Sagar, V. R., ... & Kishore, N. (2018).

Vegetable varieties with multiple attributes spread at

faster rate-A case study in popularizing carrot variety

Pusa Rudhira in NCR Region. Indian Journal of

Horticulture, 75(3), 482-485.

Putra, D. W. T., & Punggara, A. A. (2018). Comparison

analysis of simple additive weighting (SAW) and

weigthed product (WP) in decision support systems. In

MATEC Web of Conferences (Vol. 215, p. 01003). EDP

Sciences.

Surbakti, E. E., Purwandari, B., Solichah, I., &

Kumaralalita, L. (2019, September). Analysis of

software development method selection: a case of a

private financial institution. In Proceedings of the 3rd

International Conference on Business and Information

Management (pp. 168-173).

Pan, Z., Zhang, R., & Zicari, S. (Eds.). (2019). Integrated

Processing Technologies for Food and Agricultural By-

Products. Academic Press.

Sundari, M. T. (2011). Analisis biaya dan pendapatan usaha

tani wortel Di kabupaten karanganyar. SEPA: Jurnal

Sosial Ekonomi Pertanian dan Agribisnis, 7(2).

Feng, Q., He, D., Zeadally, S., Khan, M. K., & Kumar, N.

(2019). A survey on privacy protection in blockchain

system. Journal of Network and Computer

Applications, 126, 45-58.

Linden, I., Liu, S., & Colot, C. (Eds.). (2017). Decision

Support Systems VII. Data, Information and Knowledge

Visualization in Decision Support Systems: Third

International Conference, ICDSST 2017, Namur,

Belgium, May 29-31, 2017, Proceedings (Vol. 282).

Springer.

Simon, P. W. (2019). Classical and molecular carrot

breeding. In The carrot genome (pp. 137-147).

Springer, Cham.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

820