Claim Missing Document
Check
Articles

Found 30 Documents
Search

IMPLEMENTASI ARTIFICIAL INTELLIGENCE DALAM APLIKASI CHATBOT SEBAGAI HELPDESK OBJEK WISATA PANTAI DI-BATAM DENGAN METODE FORWARD CHAINING Situmorang, Jonatan; Koko Handoko
Computer Science and Industrial Engineering Vol 10 No 3 (2024): Comasie
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v10i3.8511

Abstract

The chatbot application for recommending beach destinations in Batam City utilizes artificial intelligence and the forward chaining method on IBM Cloud's Watson Assistant. In the era of advancing technology, many individuals prefer seeking information through Google. Therefore, researchers have designed a website interface and chatbot to provide information assistance. The website features main menus such as home, tourism, gallery, contact, and chatbot, facilitating users in finding information about beaches in Batam. The forward chaining method, a type of inference engine, is employed to trace rules and offer solutions based on information gathering. The research outcomes encompass the website interface and chatbot displays, subjected to testing to ensure alignment with requirements. Usecase diagrams are employed in designing interactions between users and the system. Using Watson Assistant on IBM Cloud, the chatbot provides beach recommendations with greetings and questions upon the initial display. Accessing information through this website simplifies things for the community, eliminating the need to search for information from other sources. Through the forward chaining method, chatbot responses are determined based on pre-established rules. The objective of this application is to provide an efficient and user-friendly information search experience for Batam City users.
APLIKASI EDUKASI MATA PELAJARAN ILMU PENGETAHUAN ALAM BERBASIS ANDROID Hutapea, Indra; Koko Handoko
Computer Science and Industrial Engineering Vol 10 No 4 (2024): Comasie
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v10i4.8577

Abstract

Education according to the law concerning the National Education System No. 20 of 2003 Article 37 Paragraph 1 That the primary and secondary education curriculum must contain natural science (IPA). Natural science study materials are intended to develop students' knowledge, understanding and analytical skills of the natural environment and its surroundings. The scope of science includes aspects of living things, objects, energy, earth and nature. The RAD method is a method that focuses on developing applications quickly through iterative development and feedback, in accordance with the target of a shorter application development time. At present, with increasingly high and advanced technological developments, the use of Android smartphones in education is still less visible in the activities that students tend to do on social media and games so that the available time is not utilized properly, Some schools still use the old method, namely only using books, especially regarding the use of teaching materials and implementation of learning, teaching materials have an influence on students. With the problems that exist at this location, what is offered in this research is to build a system with research results in the form of an Android-based application for natural science subjects.
ANALISIS SENTIMEN UNTUK MEMPREDIKSI PENGARUH PENGGUNAAN GADGET TERHADAP PENDIDIKAN DENGAN METODE NAIVE BAYES Hutagalung, Isnaini; Koko Handoko
Computer Science and Industrial Engineering Vol 10 No 4 (2024): Comasie
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v10i4.8582

Abstract

In this age of rapid technological development, everyone including children needs access to communication and information technology. One of the electronic devices that people use to communicate is gadgets. People also often use gadgets for business, social media, news, and human resources. It is true that we often see adults, the elderly (60 years old and above), and teenagers (12-21 years old) using multiple gadgets. More ironically, these devices are also often used and given to toddlers (1-5 years old), who should not have them yet, but are given by their parents to keep their children from crying. Therefore, we will use data mining techniques combined with the Naive Bayes approach to conduct this research. The Naive Bayes method is one of the strategies used in data mining to classify data. The results of the study are based on the amount of time students spend in using devices for learning that has been recorded and then analyzed using the RapidMiner application with the Naive Bayes method
IMPLEMENTASI DATA MINING PADA PREDIKSI PENJUALAN PRODUK TERLARIS DENGAN METODDE K-NEAREST NEIGHBOR Pratama, Khevind Adrian; Koko Handoko
Computer Science and Industrial Engineering Vol 11 No 4 (2024): Comasi Vol 11 No 4
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v11i4.9028

Abstract

Jesindo Mitra Prakarsa stores in Batam City include stores that offer a range of toys for kids. Lack of information about which items clients buy frequently or infrequently leads to an overabundance of inventory, which is a common issue for the store selling toys. Therefore, a projection that makes use of data or historical sales information is needed to assist the store in stocking the goods. The goal of this study is to use the K-Nearest Neighbor algorithm to predict the sales of the most popular toys among kids in Jesindo Partners Prakarsa stores. Data will be gathered through observation techniques, in-store interviews, and literature reviews relevant to the study's subjects. In order to forecast sales for the following month, the author used the Euclidean Distance formula with a value of k=3 and RapidMiner software to predict sales of the best-selling product, Tricycle Happy. The formula had a target of seven products, but it was predicted to sell as many as six. Based on the results of RMSE testing, which showed a value achieved close to zero at 2.035 +/- 0.000 means, the author's algorithm matched or was effectively applied to this study.
PREDIKSI PENJUALAN DISTRIBUTOR CV. LESTARI MANDIRI JAYA MENGGUNAKAN METODE FUZZY MAMDANI Seven, Sevrius Ndruru; Koko Handoko
Computer Science and Industrial Engineering Vol 11 No 2 (2024): Comasie Vol 11 No 2
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v11i2.9029

Abstract

Distribution firms frequently encounter difficulties in forecasting distributor sales due to the absence of a well-organized and precise prediction system. Inaccurate sales forecasts can result in issues that disturb business operations. The objective of this study is to identify sales data variables through fuzzy Mamdani calculations to forecast distributor sales for the company. To solve the sales prediction challenge, this research utilizes MATLAB software. Numerical data is vital in data processing as the fuzzy Mamdani logic method necessitates numerical data for processing in MATLAB. Once the data conversion process is finished, the subsequent step is to implement it in MATLAB to derive research findings that involve input and output variables. Tests were performed using two approaches: manual calculations and MATLAB. In every test, both manual and MATLAB methods were employed for defuzzification. The findings from both methods showed a sales increase with the same value of 85, indicating consistency between the methods and validating the accuracy of the implemented fuzzy model. Utilizing the fuzzy Mamdani method and MATLAB implementation allows companies to forecast distributor sales with greater precision. This analysis assists companies in recognizing sales trends and developing effective strategies to boost distributor sales.
IMPLEMENTASI DATA MINING UNTUK MEMPREDIKSI PENGARUH MEDIA SOSIAL TERHADAP SEMANGAT BELAJAR ANAK Rahayu, Shintya; Koko Handoko
Computer Science and Industrial Engineering Vol 11 No 2 (2024): Comasie Vol 11 No 2
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v11i2.9046

Abstract

In the digital era, the internet has become essential to daily life for all age groups, including children. Social media platforms like WhatsApp, YouTube, and TikTok are now integral to daily life, serving communication, news dissemination, entertainment, and promotional purposes. However, excessive use can lead to addiction and negatively impact learning, especially among children at the Al-Ikhlas Orphanage. This study employs data mining with the Naïve Bayes algorithm to analyze survey data on social media usage and its impact on learning enthusiasm. Naïve Bayes was selected for its high classification and prediction accuracy. Using RapidMiner software, the study found that social media significantly influences children's learning enthusiasm, achieving an accuracy rate of 85%. For the "strongly agree" class, precision is 92.86% and recall is 86.67%, while for the "disagree" class, precision is 66.67% and recall is 80.00%. The results indicate a significant influence of social media on children's learning enthusiasm.
PENERAPAN DATA MINING UNTUK PREDIKSI PENGARUH PENGGUNAAN APLIKASI GETCONTACT TERHADAP KEAMANAN PENGGUNA DALAM MENERIMA PANGGILAN DAN PESAN Monika, Mayana Kris; Koko Handoko
Computer Science and Industrial Engineering Vol 11 No 4 (2024): Comasi Vol 11 No 4
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33884/comasiejournal.v11i4.9088

Abstract

The study predicts the impact of the GetContact application on user security by using data mining using Naïve Bayes techniques. By using GetContact as a spam and call information filter, users can find and filter unwanted calls and messages. The app send spam or scam notifications based on the labels on the numbers it calls, with milions of users worldwide. Using RapidMiner software to analyze the data, this study used user data in the Batamindo Dormitory Block Q17 Muka Kuning. There were a total of 91 data used, divided into 61 training data and 30 test data. The result of the prediction of the influence the use of the GetContact application with Naïve Bayes resulted in model performance of 83,33% accuracy, 80% precision, and 90% recall. These result show that Naïve bayes successfully predicted the test data class. Of the 30 test data, 16 have a “Yes” class and 14 have a “No” class. This shows that 16 out of 30 users believe that the GetContact app compromises user security, while 14 out of 30 users feel the opposite.
Data Mining Pada Jumlah Penumpang Menggunakan Metode Clustering Koko Handoko; Lido Sabda Lesmana
Prosiding Vol 1 (2018): SNISTEK
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Currently, the concept of Data Mining is increasingly recognized as an important tool in information management because of the increasing amount of information. This study aims to classify the number of passengers at the airport Hang Nadim Batam. A very large sum of these huge amounts of data wants data for the new knowledge of the data. Data drilling is a process that uses statistical, mathematical, artificial intelligence, and machine learning techniques to extract and identify relevant information and large databases. One technique known in data mining is grouping, which is the process of grouping a number of data or objects into a cluster (group). Each of these clusters will contain data that is as similar as possible and different from the objects in the other clusters. In this case we can select the price data as the initial cluster center, then calculate the distance between each data in the cluster center and determine the nearest cluster, then average averages of all groups, so that the existing process is not the same. After the process is done quickly by using RapidMiner, create clusters in grouping the number of passengers. Where the variable used is the first variable, namely: the number of passengers coming, the number of passengers departing, the number of transits that transit. Where will present data on the number of passengers a lot, medium and small.
Data Mining Clustering Nilai IQ Siswa Sekolah SMAK Yos Sudarso Batam Alvendo Wahyu Aranski; Koko Handoko
Prosiding Vol 2 (2019): SNISTEK
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Sekolah Yos Sudarso Batam merupakan suatu sekolah swasta yang banyak diketahui oleh masyrakat batam, sekolah berhasil memperoleh peringkat pertama, dalam rata-rata nilai ujian nasional tingkat Kepulauan Riau. Pengelompokkan siswa berdasarkan tingkat IQ bertujuan agar siswa dapat menerima metode pembelajaran yang lebih tepat dan efektif sehingga siswa dapat menerima pelajaran dengan baik. Pada sekolah SMAK Yos Sudarso batam yang menjadi tempat penelitian, siswa yang mendaftarkan diri untuk bersekolah sangat banyak dan pengelompokkan siswa akan memakan waktu yang sangat banyak dan tidak efisien jika dilakukan secara manual. Pengelompokkan siswa dapat dilakukan secara cepat dan tepat dengan memanfaatkan data mining. Salah satu metode pada data mining adalah pengelompokkan atau clustering. Tahapan metode ini dimulai dengan pemilihan secara acak K, K disini merupakan banyaknya cluster yang ingin dibentuk. Kemudian tetapkan nilai-nilai K secara random, untuk sementara nilai tersebut menjadi pusat dari cluster atau biasa disebut dengan centroid, mean atau “means”. Hitung jarak setiap data yang ada terhadap masing-masing centroid menggunakan rumus Euclidian hingga ditemukan jarak yang paling dekat dari setiap data dengan centroid setelah selesai perhitungan secara manual di jalankan dengan software data mining yaitu RapidMiner. Penggunaan metode K-Means untuk melakukan pengelompokkan merupakan salah satu metode yang tepat jika dilihat dari variabel yang akan digunakan yaitu nilai tingkat IQ siswa.
Rancangan Pemilihan Paket Pelaminan Menggunakan Finite State Automata di Kota Batam Koko Handoko; Alvendo Wahyu Aranski
Prosiding Vol 2 (2019): SNISTEK
Publisher : LPPM Universitas Putera Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Paket pelaminan untuk pernikahan di kota Batam sangat banyak ragamnya. Hal ini membuat calon pengantin menjadi bingung dalam menentukan pilihannya. Ada sekian banyak pertimbangan yang membuat calon pengantin harus memilih satu diantara sekian banyak pilihan paket pelaminan, diantaranya harga paket, banyak baju yang akan digunakan, sewa fotografer, dan jenis tenda yang akan digunakan. Dari sekian banyak input yang diinginkan oleh calon pengantin untuk memilih paket pelaminan yang ada di kota Batam, ada baiknya hal tersebut disederhanakan prosesnya dengan bantuan Finite State Automata. Di dalam FSA, terdapat cara atau diagram yang bisa digunakan untuk membuat keputusan yang simpel dalam pengambilan keputusan. Diagram yang ditawarkan bisa membantu calon pengantin untuk menentukan pilihannya terhadap pemilihan paket pelaminan. Di dalam FSA, biasanya diawali dengan simbol awal dan diikuti dengan keadaan yang banyak sesuai dengan input yang dimiliki. Dengan menggunakan fungsi transisi pada FSA, penentuan pemilihan paket pelaminan bisa ditelusuri dengan mudah, sehingga pilihan paketnya jatuh kepada pilihan yang tepat. Pada rancangan ini, peneliti membuatkan diagram FSA dengan tujuan agar ketika diubah atau digunakan oleh alat pembuat keputusan yang lain bisa membantu dan mempermudah dalam pembuatan aturan sehingga alat yang digunakan untuk mengambil keputusan bisa menghasilkan nilai atau hasil sesuai dengan yang diharapkan. Di dalam penelitian ini, FSA yang peneliti gunakan adalah Deterministic Finite State Automata, yang mencakup kegiatan penelusuran hasil dengan cara memberikan satu input kesetiap state yang ada. Kesimpulannya, FSA dapat digunakan untuk membuat suatu rule dalam logika fuzzy sehingga mempermudah pekerjaan peneliti dalam membuatkan rule pada aplikasi matlab nantinya.