Claim Missing Document
Check
Articles

Found 5 Documents
Search

Service Quality Analysis Using Servqual and Importance Performance Analysis (IPA) Methods (Case Study of Sidomulya Village, Astanapura District) Nendi, Iksan; Nahdiyyah, Iim Rohimatun; Wiguna, Tantra Agun; Saputra, Andika Bagus; Tarsini, Iin
Journal of Social Research Vol. 4 No. 6 (2025): Journal of Social Research
Publisher : International Journal Labs

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55324/josr.v4i6.2583

Abstract

Public services at the village level play a crucial role in government implementation and significantly affect community satisfaction. However, gaps often exist between public expectations and their perceptions of service quality. This study aims to measure the quality of public services in Sidomulya Village, Astanajapura District, using the Service Quality (Servqual) method to identify the gap between expectations and perceptions, and to determine priority areas for improvement using Importance-Performance Analysis (IPA). Data were collected via questionnaires distributed to 30 respondents and analyzed through validity and reliability tests, followed by Servqual gap analysis and IPA quadrant mapping. The Servqual analysis revealed negative gaps across all dimensions of service quality—Tangibles (-0.39), Reliability (-0.97), Responsiveness (-0.53), Assurance (-0.40), and Empathy (-0.20)—with an average gap of -0.83, indicating that service quality does not meet public expectations. The largest gap was found in the Reliability dimension, highlighting concerns about service dependability. IPA results showed that aspects P1 and P9, related to administrative services and targeted social assistance, respectively, fall into the Top Priority quadrant, signaling urgent need for improvement. These findings suggest that the village government should prioritize enhancing these key service areas to better align with community expectations, thereby increasing public trust and satisfaction. This integrated application of Servqual and IPA provides a strategic framework for evaluating and improving village-level public services.
Mendeteksi Sarkasme Di Komentar Twitter Menggunakan Bilstm Dengan Penyetelan Hiperparameter Tarsini, Iin
Mutiara: Multidiciplinary Scientifict Journal Vol. 1 No. 12 (2023): Mutiara: Multidiciplinary Scientifict Journal
Publisher : Al Makki Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.57185/mutiara.v1i12.433

Abstract

Di era digital, sarkasme sering digunakan dalam komunikasi online, terutama di platform media sosial seperti Twitter. Namun, deteksi sarkasme otomatis merupakan tantangan utama dalam Natural Language Processing (NLP), karena sarkasme bergantung pada konteks yang lebih luas yang sulit diidentifikasi dengan teknik tradisional. Oleh karena itu, metode yang dapat memahami konteks lebih dalam dan mendalam diperlukan untuk meningkatkan akurasi deteksi sarkasme. Penelitian ini bertujuan untuk mengembangkan dan menguji model deteksi sarkasme menggunakan Bidirectional Long Short-Term Memory (BiLSTM) yang dioptimalkan dengan hyperparameter tuning, untuk meningkatkan kinerja deteksi sarkasme dalam teks Twitter. Penelitian ini menggunakan dua dataset utama, yaitu Twitter Sarcasm Data dan SST-2 Sarcasm Dataset, yang telah diberi label secara manual. Model BiLSTM dibangun menggunakan pustaka Keras dan TensorFlow, dan penyetelan hiperparameter dilakukan dengan Pencarian Grid dan Pencarian Acak untuk meningkatkan kinerja model. Model BiLSTM menunjukkan performa terbaik dengan akurasi 85%, akurasi 82%, dan skor F1 80,5%. Penyusunan hiperparameter yang tepat juga meningkatkan penarikan dan mengurangi overfitting dibandingkan dengan model lain, seperti LSTM, CNN, dan SVM.
Explore flowchart and pseudocode concepts in algorithms and programming Tarsini, Iin; Anggraeni, Riska
Indonesian Journal of Multidisciplinary Science Vol. 3 No. 5 (2024): Indonesian Journal of Multidisciplinary Science
Publisher : International Journal Labs

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55324/ijoms.v3i5.807

Abstract

Programming algorithms are a class of mathematical algorithms that are widely used in computer science and engineering. They are used to solve problems involving mathematical calculations, but in their implementation, algorithms do not always handle complex calculations, especially when genetic algorithms are often used. This article uses the method of literature study or literature review to evaluate several studies related to the topic discussed, especially in introducing the basic concepts of programming algorithms, especially for beginners, focusing on the basic understanding required. The research includes a review of algorithmic concepts, algorithm notation, characteristics, properties, and the basic structure of algorithm learning, along with the steps of implementing programming algorithms. The conclusion of this article is that learning basic programming algorithms is an essential necessity for anyone who wants to get started in the world of programming.
Optimization of Natural Resources to Realize Community Welfare Kholipah, Siti Ainul; Zaenal Asikin, Muhamad; Mijoyo, Mijoyo; Tarsini, Iin; Khoerudin, Muhamad
Jurnal Indonesia Sosial Teknologi Vol. 5 No. 2 (2024): Jurnal Indonesia Sosial Teknologi
Publisher : Publikasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59141/jist.v5i2.910

Abstract

The purpose of this study is to empower the people of Panambangan Village in realizing independence and welfare based on optimizing community potential, focusing on understanding the history of Panambangan Village, the origin of the name of the village, and the potential of its natural resources. The method used in this study is qualitative research method. The potential of Panambangan Village tourist attraction has a tourist attraction that until now has not been developed. Research results The wealth of tourism potential of Panambangan Village needs to be developed in order to support sustainable tourism. Recommendations that can be given to develop tourism potential are directed based on the condition of Panambangan Village whose status as a tourist village, but conditions on the ground have not reflected this. This is shown by the development of tourism which is still constrained by development costs. Even though Panambangan Village has many natural features and cultural traditions that can be considered quite exciting tourist attractions, The conclusion of this study shows that efforts to optimize natural resources can be vital in achieving community welfare by combining ecological sustainability and economic development.
Pengembangan Model Cerdas Monitoring Tanaman dan Rekomendasi Pemupukan Presisi Berbasis IoT dan Machine Learning di Kabupaten Cirebon Sudrajat, Sudrajat; Khoerudin, Muhammad; Toto, Toto; Tarsini, Iin; Hauzaan, Mohamad Hisyam
Jurnal Locus Penelitian dan Pengabdian Vol. 4 No. 11 (2025): JURNAL LOCUS: Penelitian dan Pengabdian
Publisher : Riviera Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58344/locus.v4i11.5050

Abstract

Pertanian presisi menuntut kestabilan larutan nutrisi untuk mendukung pertumbuhan optimal hidroponik. Penelitian ini mengembangkan sistem IoT cerdas untuk monitoring dan pengendalian otomatis pH–TDS serta rekomendasi pemupukan cair pada pakcoy (Brassica rapa L.) di Greenhouse Kamarang–Tandang, Cirebon. Penelitian ini bertujuan mengembangkan sistem cerdas berbasis IoT dan Machine Learning (ML) untuk monitoring dan kontrol otomatis parameter nutrisi serta memberikan rekomendasi pemupukan presisi. Metode yang digunakan adalah quasi-eksperimental dengan membandingkan kontrol Manual, Ambang, Fuzzy-only, dan Hybrid Fuzzy-ML. Sistem memadukan sensor pH, TDS/EC, suhu–kelembapan dan ESP32 yang mengendalikan pompa peristaltik (AB-mix, pH-up, pH-down). Kendali fuzzy menerapkan strategi pH-first ? jeda homogenisasi ? TDS dengan micro-dosing bertahap, sementara Machine Learning (Ridge/Random Forest/XGBoost) berfungsi sebagai look-ahead untuk memprediksi drift 30–60 menit ke depan dan menyarankan koreksi kecil di bawah guardrail keselamatan. Telemetri harian (pH, TDS, suhu air/udara, RH, aksi pompa) digunakan untuk evaluasi proses dan pelatihan model walk-forward. Sistem berhasil menjaga pH mendekati target (?6,0) dan TDS dalam rentang 800–1200 ppm dengan osilasi lebih rendah serta respons pascakoreksi lebih halus dibanding operasi manual. Model Ridge menurunkan kesalahan prediksi TDS dari ±127 ppm menjadi ±28 ppm, dan Random Forest memperbaiki kesalahan pH dari ±0,95 menjadi ±0,87 unit, memungkinkan penjadwalan koreksi dini dan kecil. Secara agronomis, stabilitas kimia ini selaras dengan peningkatan konsistensi tinggi dan jumlah daun selama fase vegetatif. Kontribusi utama riset adalah hibridisasi fuzzy–ML yang tetap aman (audit log, rollback), mudah direplikasi ke komoditas lain melalui penyesuaian set-point, parameter kimia, dan retraining model. Temuan ini menegaskan potensi implementasi pertanian presisi berbasis IoT–AI pada skala greenhouse lokal dan menjadi dasar pengembangan uji A/B serta ekspansi ke lahan terbuka yang kompatibel.