<![CDATA[Abstrak. Penelitian ini bertujuan untuk menganalisis pengaruh kualitas pelayanan PT. GO-JEK Indonesia terhadap tingkat kepuasan konsumen di kalangan mahasiswa Universitas Pamulang menggunakan Fuzzy Inference System (FIS) Tsukamoto. Sampel dalam penelitian ini berjumlah 150 responden yang terbagi rata ke dalam tiga kelompok shift perkuliahan, yaitu Reguler A, Reguler B, dan Reguler C, masing-masing sebanyak 50 responden dengan proporsi gender berimbang (25 pria dan 25 wanita). Pengukuran kualitas pelayanan menggunakan lima dimensi utama, yaitu Bukti Fisik (Tangibles), Kehandalan (Reliability), Daya Tanggap (Responsiveness), Jaminan (Assurance), dan Empati (Empathy). Hasil uji validitas menunjukkan seluruh butir pernyataan kuesioner valid (r xy > 0,159) dan reliabilitas instrumen sangat tinggi (r11 = 0,859). Analisis data dilakukan dengan konversi skala Likert dan pemodelan aturan logika fuzzy pada perangkat lunak Matlab. Melalui proses defuzzifikasi metode rata-rata terbobot (weighted average), diperoleh nilai akhir kepuasan sebesar 80,93%. Angka ini berada pada skala interval 75%–100%, yang berarti bahwa para pengguna di lingkungan akademis merasa sangat puas terhadap kualitas layanan transportasi ojek online yang diberikan. Pemodelan ini memberikan akurasi penegasan yang lebih adaptif dalam merepresentasikan ketidakpastian persepsi kepuasan manusia dibandingkan metode konvensional. Abstract. This study aims to analyze the effect of service quality provided by PT. GO-JEK Indonesia on customer satisfaction among students at Pamulang University using the Tsukamoto Fuzzy Inference System (FIS). The sample consisted of 150 student respondents divided equally into three lecture shifts (Regular A, Regular B, and Regular C), with 50 respondents each and a balanced gender proportion (25 males and 25 females). Service quality measurement uses five main dimensions: Tangibles, Reliability, Responsiveness, Assurance, and Empathy. The validation test results showed that all 15 questionnaire items were valid (r xy > 0.159) and the instrument reliability was very high (r11 = 0.859). Data analysis was performed using Likert scale conversion and fuzzy logic rule modeling via Matlab software. Through the defuzzification process using the weighted average method, a final satisfaction score of 80.93% was obtained. This value falls within the 75%–100% interval scale, indicating that users in the university environment are highly satisfied with the online motorcycle taxi transportation service. This modeling provides a more adaptive resolution in representing the uncertainty of human satisfaction perception compared to conventional methods.]]>
