JDMIS: Journal of Data Mining and Information Systems

Evaluasi Kinerja Bisnis Berbasis Business Intelligence Dashboard Pada UD. Sentral

2025-04-15T06:04:21+00:00

This study aims to evaluate the effectiveness of implementing a Business Intelligence (BI) Dashboard in supporting the business performance evaluation process at UD. Sentral, a trading company engaged in the distribution of data cards and voucher products. The main issues identified include manual data recording and reporting processes that are time-consuming, prone to errors, and lack informative visualizations for management. A qualitative approach was used through interviews and direct observation with the business owner and operational staff to identify user needs and challenges. Operational data were then processed and visualized using Power BI in the form of a dashboard displaying Key Performance Indicators (KPIs), such as total sales, stock movement, customer count, and monthly sales trends. The implementation results showed that the BI dashboard improved business monitoring efficiency, accelerated evaluation processes, and supported strategic data-driven decision-making. Furthermore, the dashboard encouraged a data-driven culture and enabled early identification of declining business performance. These findings contribute to the development of data-based information systems for MSMEs in Indonesia.

Analisis Sentimen Ulasan Perbedaan Aplikasi BCA Mobile dengan MYBCA di Playstore Menggunakan Metode Lexicon

2025-04-24T11:30:13+00:00

This study aimed to analyze and compare user sentiment toward two digital banking applications owned by Bank Central Asia: BCA Mobile and myBCA, based on user reviews collected from the Google Playstore. The research employed a text-based quantitative approach, using a lexicon-based sentiment analysis method to classify user opinions into positive, negative, and neutral categories. Data were collected through web scraping and processed using text preprocessing techniques such as tokenization, stopword removal, and stemming. The results showed that most reviews were neutral, with myBCA receiving a higher proportion of positive reviews than BCA Mobile. Visual analyses through charts and wordclouds successfully illustrated differences in user perception related to application features such as login convenience, user interface design, and technical issues. This study concluded that sentiment analysis is an effective tool to evaluate user experience and provide strategic insights for the future development of digital banking services.

Analsis Sentimen Churn Pelanggan dalam Layanan Streaming NETFLIX di X Menggunakan Metode IndoBERT

2025-05-05T06:14:23+00:00

This study aims to analyze customer sentiment toward Netflix’s streaming service as expressed on social media platform X (formerly Twitter), in order to identify potential churn. The research employs a combination of Text Mining and Sentiment Analysis methods, utilizing the IndoBERT-based Natural Language Processing (NLP) model. Data was collected using web scraping techniques with keywords indicating complaints or cancellation of Netflix subscriptions. The text data underwent preprocessing steps including case folding, cleaning, lemmatization, and tokenization. Sentiment classification results showed that most tweets expressed negative sentiment, suggesting a high risk of customer churn. Key factors driving negative sentiment include subscription pricing, login policy restrictions, and the cancellation of popular content. These findings can assist Netflix’s marketing and product development teams in creating data-driven retention strategies. Furthermore, the study demonstrates that the IndoBERT model is effective in classifying Indonesian-language social media opinions into positive, neutral, and negative sentiment categories.

Penerapan Algoritma K-Nearest Neighbors untuk Klasifikasi Kualitas Air Minum

2025-05-28T03:40:34+00:00

This study aims to classify drinking water potability based on physical and chemical parameters using the K-Nearest Neighbors (KNN) algorithm. The dataset, sourced from the Kaggle platform, contains 100,000 water samples with nine key attributes, including pH, hardness, total dissolved solids (TDS), sulfate, chloramines, conductivity, organic carbon, trihalomethanes, and turbidity. The target label is potability, indicating whether the water is safe (1) or unsafe (0) for consumption. The preprocessing steps included normalization and splitting the data into training and testing sets. The KNN model was trained by experimenting with various K values to achieve optimal performance. Evaluation using a confusion matrix showed that the model achieved an accuracy of 78%. For the potable class, the model reached a precision of 72%, recall of 91%, and F1-score of 81%. For the non-potable class, it achieved a precision of 88%, recall of 65%, and F1-score of 75%. Although the model tends to misclassify unsafe water as safe, overall performance is promising. These findings suggest that the KNN algorithm can serve as an effective classification approach and has potential for application in automated water quality monitoring systems.

Penilaian Kualitas Layanan WiFi Oxygen dan Kolerasinya terhadap Kepuasan Pengguna

2025-06-06T10:54:14+00:00

Penelitian ini merupakan penelitian yang dilakukan untuk mengevaluasi pengaruh kualitas layanan WiFi Oxygen terhadap tingkat kepuasan pengguna berdasarkan beberapa parameter teknis jaringan yang diuji. Data yang digunakan berasal dari log sistem monitoring internal dan survei kepuasan pengguna dengan jumlah responden sebanyak 198 orang. Metode yang digunakan dalam penelitian ini adalah regresi linear berganda yang digunakan untuk mengidentifikasi pengaruh variabel kualitas jaringan seperti kecepatan download, kecepatan upload, latency, packet loss, dan jitter terhadap kepuasan pengguna. Hasil penelitiannya menunjukkan bahwa kualitas jaringan tidak terlalu berpengaruh signifikan terhadap kepuasan berdasarkan kecepatan jaringan dengan nilai R Square sebesar 11.6%. Sebaliknya, model regresi untuk kepuasan pengguna berdasarkan kestabilitasan jaringan lebih baik dengan nilai R Square sebesar 36.6%, di mana variabel kecepatan download dan jitter berpengaruh positif dan signifikan. Temuan ini menunjukkan bahwa kestabilitasan jaringan lebih berkontribusi pada kepuasan pengguna dibandingkan kecepatan semata. Penelitian ini memberikan dasar untuk perbaikan layanan WiFi dengan fokus pada peningkatan kestabilitasan jaringan.

Implementasi Market Basket Analysis Dengan Algoritma Frequent Pattern Growth Pada Data Transaksional di Electronic Commerce

2025-06-09T12:16:26+00:00

The Growth of the e-commerce industry has resulted in a massive volume of transaction data, necessitating effective data analysis techniques to extract customer purchasing patterns. The Frequent Pattern Growth (FP-Growth) algorithm is one of the data mining methods that can be used to identify frequently occurring purchase patterns without explicitly generating candidate itemsets. This study aims to implement and evaluate the performance of the FP-Growth algorithm in analyzing e-commerce transaction data to identify recurring shopping patterns. The research methodology includes transaction data collection, data preprocessing, FP-Growth algorithm implementation, and result analysis. This study utilizes an e-commerce transaction dataset from an online retail store based in the United Kingdom, comprising 541,909 transaction records. The research findings indicate that the FP-Growth algorithm is efficient in identifying frequently occurring transaction patterns. Using a support threshold of 1% and a confidence level of 80%, 13 association rules were discovered, demonstrating relationships between frequently co-purchased products. Further analysis shows that these findings can be leveraged by e-commerce businesses to develop marketing strategies based on product recommendations. In conclusion, the FP-Growth algorithm is an effective approach for extracting purchasing patterns from large-scale e-commerce transaction data.

Deteksi Sentimen Komentar Aplikasi Gobis Suroboyo dengan Metode Naive Bayes dan Metode Regresi Logistik

2025-06-13T02:45:49+00:00

This research discusses sentiment analysis of user comments on the Gobis Suroboyo application using the Naive Bayes algorithm and Logistic Regression. Data was obtained through web scraping method from Google Play Store, with a total of 1,015 comments which then went through text pre-processing such as data cleaning, case folding, stemming, normalisation, filtering, tokenizing, and feature selection using TF-IDF. Sentiment labels were determined based on user ratings, with ratings above 3 as positive and 3 and below as negative. The results show that the Naive Bayes algorithm is better at classifying positive sentiment with a precision of 81% and f1-score of 77%, while Logistic Regression excels at negative sentiment with a precision of 82% and f1-score of 82%. The WordCloud visualisation shows dominant words such as “app”, “good”, and “bus stop” that reflect users attention to the app features and transportation services. The findings show that both algorithms have competitive and reliable performance for evaluating public opinion on comment-based digital services. This research is expected to be a reference for app developers and local governments in improving the quality of digital public services.