• Tidak ada hasil yang ditemukan

Sadar Putra Sadeli Ari Purno - Jurnal

N/A
N/A
Protected

Academic year: 2023

Membagikan "Sadar Putra Sadeli Ari Purno - Jurnal"

Copied!
8
0
0

Teks penuh

(1)

675

JURNAL DARMA AGUNG, Vol. 30, No. 1, (2022) April : 675 - 682

SISTEM INFORMASI REKOMENDASI PRODUK PADA WEBSITE ECOMMERCE DENGAN MENGGUNAKAN ALGORITMA APRIORI

By:

Sadar Putra Sadeli 1) Ari Purno Wahyu Wibowo 2) Universitas Widyatama, Bandung 1,2)

E-mail :

sadar@widyatama.ac.id1) ari.purno@widyatama.ac.id 2)

ABSTRACT

The current e-commerce sales system makes it easier for someone to open a business place or shop without having to need a lot of places with even more catalogs of goods. The problem faced today is if the catalog of goods sold is very large and it is very difficult to determine the number of items that are most sold and the categories of favorite products, then an e-commerce system is created using an a priori algorithm that can determine the most categories of goods. sold by determining the number of rules or data for each transaction, this system can also read purchase patterns made by consumers so that each consumer's purchase data can be read and will later be used to determine product promotions for each consumer. Determining a product recommendation can later be used to determine a marketing strategy, the Apriori algorithm has a simple formula, namely by determining the value of support and confidence, support means the number of items purchased the most and confident is the calculation of an estimate with a percent value of the number of items purchased. The apirori algorithm can easily and precisely see the pattern of purchases of goods made by consumers and by category these goods can be displayed on the website's main page. These favorite products can be used as marketing media to promote the most purchased items as well as determine the stock of goods.

Keyword: E-Commerce; Algoritma Apriori

1. INRODUCTION

Currently internet technology is developing rapidly, we can see this with the increasing number of users using internet facilities not only to get the latest news and information needed and to connect with other people in cyberspace, but the internet has also become a medium of commerce. between companies and

business units and consumers. Trading through internet technology or e- commerce (electronic commerce) offers enormous market opportunities. It's very expensive to miss. Especially for companies or business units that want to market (sell) their products globally and not only in certain areas. Transactions that are still carried out manually, namely by

(2)

676

telephone or directly to certain shops, of course, this method is quite an obstacle for overseas buyers/consumers. Selling products online generates greater profits than selling products through traditional methods/means such as telephone, fax, or direct to point of sale.

Having Components A system consists of a number of components that interact with each other, working together to form a unit. System components can be a subsystem or parts of the system. A system, no matter how small, always contains components or subsystems. Each subsystem has the properties of the system to carry out certain functions and affect the overall system process. A system can have a larger system called a supra system, for example a company can be called a system and an industry which is a larger system can be called a supra system. If the industry is considered as a system, then the company can be referred to as a subsystem. Likewise, if the company is seen as a system, then the accounting system is its subsystem.

Web servers are a number of computer devices that are enabled for HTTP (Hypertext Transfer Protocol) services.

Several computers can be used as web servers by installing web software such as APACHE. Apache is a web server

software that has an open source license and is currently the most widely used, this is because Apache has advantages in terms of speed, performance and is free. As for the advantages of the Apache web server.

2. LITERATUR REVIEW

An on-line system is a system that receives input directly in the input area, records the input, and produces output that can be the result of a calculation in the area that needs it. The area itself can be broken down into orders of magnitude, such as hundreds of kilometers. Usually used as an air force reservation, railroad reservation, banking and others.

The real-time system is a control mechanism. Data recording, very fast processing, so the resulting output can be received in relatively the same time. The difference from the online system is that the time unit used is real time. Usually it's hundredths or thousandths of a second, online it's still seconds or sometimes even a few minutes. Another difference is that online usually only contains interactions with users, while real-time interactions with users and the environment being mapped.

PHP, which stands for PHP Hypertext Preprocessor, is a server-side language designed specifically for web applications.

(3)

677

JURNAL DARMA AGUNG, Vol. 30, No. 1, (2022) April : 675 - 682

PHP can be inserted between HTML languages, since it is a server side language, PHP language is executed on the server, so what is sent to the browser is

"results so" in html and php code is not visible (kadir, 2001). PHP was developed by Rasmus Lerdorf, a Unix and Perl programmer. Originally, PHP was only used to count the number of visitors who viewed Rasmus Lerdorf's CV. This script is then packaged into a tool called Personal Home Page, this package is the precursor to PHP. In September-October 1995, Rasmus created PHP/F1 version 2.0. In this version, programmers can insert structured code into HTML tags, and this code can also communicate with databases and perform complex calculations on the fly. On June 6, 1998, PHP version 3.0 was successfully released.

MySQL is a relational database management system (RDMS) distributed free of charge under the General Public License (GPL). Anyone is free to use it, but it may not be used as a commercial derivative product (Sutarman, 2003).

MySQL is a derivative of one of the main concepts in database, namely the Structured Query Language (SQL). SQL is a database operation concept, specifically for selecting and entering data, that enables automated data operations. As a database server, MySQL is superior to

other database servers when it comes to querying data. This can be proven by the speed of MySQL, which can be 10 times faster than PostgreSQL and 5 times faster than Interbase.

The Unified Modeling Language (UML) is a "language" that has become the industry standard for visualizing, designing, and documenting software systems. UML provides a standard for modeling a system. Using UML we can create models for all kinds of software applications, these applications run on any hardware, operating system, network and are written in any programming language.

However, because UML also uses classes and operations in its basic concept, it is more suitable for writing software in object-oriented languages such as C++, Java, C# or VB.NET. However, UML can still be used to model procedural applications in VB and C.

The history of UML itself is quite long, in the 1990s as we know dozens of object- oriented modeling methods have sprung up all over the world. Among them are:

Booch Methodology, Coad Methodology, OOSE Methodology, OMT Methodology, Shlaer-Mellor Methodology, Wirfs-Brock Methodology and others. This period is known as the methodological wars in object-oriented design. Each methodology carries its own notation, which introduces

(4)

678

new problems when we work with other groups/companies using other methodologies. Beginning in October 1994, Booch, Rumbaugh, and Jacobson, three men whose methodology can be described as the most widely used, pioneered object-oriented design, producing the first draft of the UML which was published in 1995.

Book search recommendation system with association rule exploration using an a priori algorithm. Borrowing books from the library and archives of East Java Province. The research steps consisted of finding data on book loans from the East Java Province Library and Archives Agency, processing the data according to a book search recommendation system that includes book codes and book names, and calculating and creating rules with a priori that were previously minimal Determining support and minimum trust, building a system with a priori application, testing by graphing support and trust execution time.

The results of this study are in the form of associations between books as a basis for alternative book loan recommendations

derived from Minsup and Minconf calculations (with data using up to 63% of transactional data for 2009-2012).

The result of the research is the formation of a decision support system that can determine the prediction of drug purchasing patterns in the Musi pharmacy.

The downside is that it does not fully explain how the system is made. The advantage is that new knowledge can be obtained based on the results of the calculation of the a priori algorithm, and the constructed system can be closely arranged in the layout of goods to make it easier

The association rule method using the a priori algorithm can find item combinations in sales transactions. The merging of articles resulting from the a priori process then becomes material for recommendations for sales promotion strategies in the form of shopping packages. The results of this research are available in the form of an information system that leads to the formation of a purchase package

(5)

679

JURNAL DARMA AGUNG, Vol. 30, No. 1, (2022) April : 675 - 682

Id_kategori Int(5) Nama _kategori Varchar(100) Kategori_seo Varchar(100) kategori

`id_orders` int(5)

`nama_kustomer` varchar(100)

`alamat` text `telpon`

varchar(20) `email` varchar(50) `status_order` varchar(50)

`tgl_order` date `jam_order` time `id_kota` int(3)

order

`id_orders` int(5) `id_produk` int(5) Order_detail

`id_orders_temp` int(5) ` id_produk` int(5) `id_session` varchar(100) `jumlah` int(5)

`tgl_order_temp` date

`jam_order_temp` time

`stok_temp`int(5) Order_temp

`id_produk` int(5)

`id_kategori` int(5) `nama_produk`) varchar(100

`produk_seo` varchar(100)

`deskripsi` text

`harga` int(20)

`stok ` int(5)

`berat` decimal(5,2) `tgl_masuk` date `gambar` varchar(100) `dibeli` int(5)

produk

Figure 1.1 table relations In the picture above is the design of a

product recommendation system table relation, this table consists of four main tables, namely the order table which is used to record the number of ordered items as well as data that gives all the number of existing stock items, in the detailed order table which explains a detail of the product which will be purchased from the origin of

the product, product manufacturer, price and stock of the product, while the category table is a menu that is used to record the grouping of products made, while the last table, namely the temporary table, is a table to view temporary ordering data and will canceled if the product is not purchased by the customer.

Beranda Keranjang Cara Pembelian Download katalog

Belanja

Produk Profil

Hubungi Kami

Banner Fun – Fun

Beranda Keranjang Cara Pembelian Download katalog Belanja

Produk

Profil Hubungi Kami

Banner

Teks keterangan profil Perusahaan

(a) Login page design (a) Product page design ( c ) Product category page design

In the picture above is a process of designing the main page of the website which has two design stages, the first

design is used by the user which displays data from the products being sold, on the main page displays the homepage data,

(6)

680

namely the main page, the profile of the website and the products presented, while the shopping cart page displays data on the

number of products purchased, and the next page is a page from the website that describes the data and the main web page.

( e ) home page implementation

( f ) Product page ( g ) Product Category Pages

In the picture above is an implementation of the main web page that displays the company logo as well as the product and website explanation profile, while in the picture (f) is the product page that displays the website's full address data, the display provides product information and goods orders, on that page the algorithm a priori provides an overview of the most sold products or recommended items, this system is processed using the a priori algorithm by processing data in the database. This recommendation system will display a data menu automatically by calculating support

and confidence from the frequency of the data obtained.

The value of support and confidence settings is set automatically by looking at the value limit of 40 support and 60 confident, on the last page, the product category will always appear on the main page menu which will be used as a reference for a product recommendation offered to customers, the system will be updated regularly automatically by comparing a transaction that will be carried out every day by comparing the amount of incoming data.

Figure 1.2 Apriori configuration

(7)

681

JURNAL DARMA AGUNG, Vol. 30, No. 1, (2022) April : 675 - 682

In the menu above is the configuration and optimization of the a priori algorithm using the PHP programming language, in this section a minimum number of supports is determined up to 100 and a

minimum confident of 100, the process is carried out so that the number of rules or data relations purchased becomes more

numerous and varied.

Table 1.1 a priori process stages.

In table 1 is a process of determining the number of confident with the calculation value supp = 5 ; and confident

= 75 ; this data can be recapitulated from previous sales data, in this data using the two data parameters "bread" and "milk" as a trial system and calculating the number of rules from the total transaction data.

3. CONCLUSION

From the results of the trial and implementation it can be concluded that the implementation of a web using the Apriori algorithm can be used and implemented, the system can automatically read goods sales data by paying attention to the total number of transactions which results from the number of rules, calculations of support and confidence will determine the number of rules, rules is a

relation of sales patterns that are carried out randomly taken from goods sales data, the function of the product recommendation-based e-commerce system can be used as a determination of stock of goods or as a reference for promotional strategies, this system makes it very easy for users to specify catalogs of goods if goods are sold in very large quantities.

4. REFERENSI

G. Ilieva ; T. Yankova, S. Klisarova.

(2015). BIG DATA BASED

SYSTEM MODEL OF

ELECTRONIC COMMERCE . Trakia Journal of Sciences , 407- 413.

(8)

682

Kasmi; Adi Nurdian Candra. (2017).

PENERAPAN E-COMMERCE

BERBASIS BUSINESS TO

CONSUMERS UNTUK

MENINGKATAN PENJUALAN PRODUK MAKANAN RINGAN KHAS PRINGSEWU. Jurnal Aktual STIE Trisna Negara ISSN : 1693-1688 , 109-116.

Nana Saharna ; Kasman Rukun . (2019).

ERANCANGAN SISTEM E-

COMMERCE BERBASIS WEB PADA TOKO INDAH SURYA FURNITURE. Jurnal Vokasional Teknik Elektronika dan Informatika .

Nur Indah Riwajanti ; Kartika Dewi Sri Susilowati ; Bimo Setyawan.

(2022). DEVELOPMENT OF ACCOUNTING INFORMATION SYSTEM BASED ON BUSINESS PROCESS MODELLING AND

NOTATION AND WEB-BASED FINANCIAL REPORT FOR MSMEs. International Journal of Economics, Business and Accounting Research (IJEBAR).

Omid H. Jader, Subhi R. M. Zeebaree, Rizgar R. Zebari. (2019). A State Of Art Survey For Web Server Performance Measurement And Load Balancing Mechanisms . INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH .

Reza Hermiati, Asnawati, . (2021).

PEMBUATAN E-COMMERCE

PADA RAJA KOMPUTER

MENGGUNAKAN BAHASA

PEMROGRAMAN PHP DAN

DATABASE MYSQL. Jurnal Media Infotama .

Referensi

Dokumen terkait

FINAL GRADE /0 Implementation Zachman Framework for Design and Analysis Commercial Web GRADEMARK REPORT GENERAL COMMENTS Instructor PAGE 1 PAGE 2 PAGE 3 PAGE 4 PAGE