DSS for Keyboard Mechanical Selection Using AHP and Profile Matching Method

Abstrak Keyboard mekanik dirancang dengan berbagai bentuk, variasi, dan spesifikasi yang berbeda dengan jenis keyboard lainnya. Keyboard mekanik sendiri memiliki fungsi estetik yang memungkinkan penggunanya untuk melakukan kustomisasi. Adanya berbagai spesifikasi pada keyboard mekanik, menyebabkan munculnya berbagai pertimbangan, yang dapat menyulitkan pengguna dalam memilih keyboard mekanik yang sesuai dengan kriteria yang diinginkan. Didukung dengan hasil pengamatan dalam Indonesia Mechanical Keyboard Group (IMKG), beberapa pengguna masih terbatas pengetahuannya mengenai produk keyboard mekanik yang tersedia di Indonesia, juga, saat ini belum terdapat solusi yang dapat menangani permasalahan tersebut. Diangkat dari permasalahan tersebut, pada penelitian ini dibangun sebuah SPK yang dapat membantu mengatasi masalah tersebut, dengan memberikan rekomendasi keyboard mekanik yang sesuai dengan keinginan pengguna. SPK diimplementasikan dalam bentuk web menggunakan metode AHP untuk proses pembobotan dan Profile Matching untuk proses skoring. Kriteria yang digunakan ditentukan dengan melakukan survei mengenai spesifikasi yang menjadi prioritas pertimbangan dalam memilih keyboard mekanik. Di akhir penelitian, SPK yang berhasil dibangun mampu memberikan rekomendasi prioritas keyboard mekanik yang sesuai dengan preferensi pengguna dan mendapatkan rata-rata hasil evaluasi sebesar 36.17 dari total nilai maksimal 40.


INTRODUCTION
Keyboard is one of the computer hardware that has an important role in providing input. With the keyboard, users can enter characters and functions into the computer system by simply pressing a button. Some keyboards not only act as input devices, but also have aesthetic and customization functions. One type of keyboard that supports this function is a mechanical keyboard.
Mechanical keyboards are designed with various shapes, variations, and specifications that are different from other types of keyboards. The mechanical keyboard itself has various specifications and variations and has an aesthetic function that allows users to customize it. With so many different specifications and variations, various considerations arise in determining the choice of a mechanical keyboard that suits you [1].
From these problems, the existence of a mechanical keyboard selection decision support system can help provide mechanical keyboard recommendations that are in accordance with the wishes of the user. The mechanical keyboard selection decision support system can assist users in finding mechanical keyboard options, by displaying product recommendations that match the criteria desired by the user.
The system is implemented in web form using the AHP weighting method and Profile Matching in decision making. The AHP weighting method is used to assist the weighting process, while the Profile Matching method is used to assess the criteria that are close to the ideal value desired by the decision maker.
The AHP method was chosen because it is a multi-criteria decision-making technique in which decision makers set priorities and determine decisions by making pairwise comparisons between criteria to get priorities in each hierarchy (saaty, 1987) [2]. While the Profile Matching method was chosen because there is an ideal level of predictor variables in each available alternative, not a minimum level that must be passed (Kusrini, 2007) [3]. Therefore, this final project discusses the development of a mechanical keyboard selection decision support system using the AHP weighting method and Profile Matching.

Research Description
In this study, the system was built using the AHP weighting method for weighting and Profile Matching for the scoring process on criteria that require user preferences. The flow of using the method on the system is shown.
First, the user provides AHP matrix input, then inputs the desired mechanical keyboard target value / preference criteria. The system will then process the input value from the user, and then display the appropriate keyboard recommendation results. For more details, the flow of using the method is shown in Figure 1

AHP Method
The Analytic Hierarchy Process (AHP) is a multi-criteria decision-making technique that aims to deal with rational and intuitive problems in determining the best alternative among a number of alternatives [4]. Comparisons can be based on actual measurements, or on a fundamental scale that shows the strength of the relative preferences between elements. The following is the Saaty fundamental scale shown in Table 3 [4]. One criteria is slightly more important. 5 One criteria is more important. 7 One criteria is strongly more important. 9 One criteria is absolutely more important. 2,4,6,8 Values between the two adjacent judgements.

Reciprocal
If criterion i is assigned with one of the value above when compared to criterion j, then criterion j has the reciprocal value when compared to criterion i.
In the AHP method, the first step is to make a pairwise comparison matrix based on the fundamental scale of the time.
(1) The next step is to normalize the pairwise comparison matrix using the following equation. (2) (3) Next, determine the priority weight by adding up the value of each row, then dividing by the number of elements. (4) After getting the AHP weights, the next step is to check the consistency of the weights (Consistency Ratio / CR). To get the CR value, first calculate the WSV value by adding up the multiplication of each row in the unnormalized pairwise comparison matrix with the priority weights of each element concerned. Then calculate the Consistency Vector / CV value by dividing the WSV by the priority weight of the appropriate element. (7) Then calculate the maximum eigenvalue ( ) by dividing the number of CV by the number of elements. (8) Then calculate the Consistency Index value, and with the following equation. (9) Then calculate the Consistency Ratio with the following equation. (10) RC stands for Random Consistency, where its value depends on the size of the matrix, or the number of elements being compared. RC values can be seen in Table 2  The AHP weight is considered consistent if it has a CR value of less than or equal to 10%, if it is more then it is considered inconsistent.

Profile Matching Method
Profile Matching is a decision-making mechanism by assuming that there is an ideal level of predictor variables that applicants must have, not a minimum level that must be passed [2]. The assessment process in the Profile Matching method can be done by assigning a direct value to the desired target, or by calculating the gap (difference between data and target). The smaller the gap, the higher the value. The steps in the Profile Matching method using the value gap are as follows: 1. Calculating the Gap value by finding the difference between the attribute/data value and the target value.

Gap = atribut valuetarget value
2. Give weight to the Gap value that has been obtained on each criterion with the values contained in Table 3 below.

Interpolasi Linear
Linear Interpolation is to determine the points between two points using a straight line function approach. In determining the Linear Interpolation equation, it can be done through a straight line equation that passes through two points P1 (X0, Y0) and P2 (X1, Y1) can be written as [5]: (12) So that the equation of Linear Interpolation is obtained as follows: (13) Description: y = point value to be searched y1 = upper limit y0 = lower limit x1 = upper limit of range x0 = lower limit of range

Dataset
The data in this study are keyboard data available in Indonesia taken from the Indonesian mechanical keyboard discussion forum on Facebook called the Indonesia Mechanical Keyboard Group. Incomplete data must be completed first by taking information from online buying and selling sites and the official keyboard brand website. Mechanical keyboard data and its specifications are stored in hard code into a table as shown in Figure 2 below.

Processing Results with AHP weighting, Profile Matching, and Linear Interpolation
After getting input from the user, the system then performs a weight calculation process using the AHP weighting method, followed by a scoring/assessment process using the Profile Matching method for criteria that require preference, and for criteria without preferences, the criteria grouping process is carried out as a cost or as a benefit. scoring is done using linear interpolation according to the type of criteria.

System Testing Results
System testing was carried out by conducting a survey involving 10 respondents from a Facebook forum/discussion group called the Indonesia Mechanical Keyboard Group. respondents were asked to try the system, then fill out the assessment form that has been provided related to the evaluation of the decision support system that has been built.
is the highest score for the statement, while is the lowest score for the statement. From the test results, obtained values for aspects of reliability 37, usability 35.5, and helpfulness 36, all of these results can be categorized into the very good category, which means the system can assist users in determining the choice of the desired mechanical keyboard. From the results of these scores, the average evaluation results for aspects of reliability, usability, and helpfulness are 36.17 out of a total maximum score of 40.