This review explores the application of machine learning (ML) in optical fiber sensing, identifying research trends and proposing a future research agenda. The study conducts a bibliometric analysis using the PRISMA 2020 framework to examine publications from Scopus and Web of Science databases. The analysis reveals that deep learning techniques and fiber Bragg gratings are extensively researched in infrastructure, particularly for structural health monitoring. However, there is a lack of research on novel materials like graphite for fiber optic sensor design. The study highlights the growing interest in ML applications for fiber optic sensors, with significant advancements in healthcare, structural monitoring, and environmental sensing. Key findings include the use of ML for detecting structural damage, predicting pipeline degradation, and improving sensor accuracy. The research also identifies emerging trends such as the integration of ML with fiber optic sensors for real-time monitoring, fault detection, and predictive maintenance. The study emphasizes the importance of interdisciplinary collaboration and the need for further research on novel materials and advanced ML techniques to enhance the performance and applicability of fiber optic sensors. The research agenda includes exploring new materials, improving sensor design, and developing more efficient ML algorithms for fiber optic sensing applications. The study concludes that the integration of ML with fiber optic sensors holds great potential for advancing various fields, including healthcare, civil engineering, and industrial monitoring.This review explores the application of machine learning (ML) in optical fiber sensing, identifying research trends and proposing a future research agenda. The study conducts a bibliometric analysis using the PRISMA 2020 framework to examine publications from Scopus and Web of Science databases. The analysis reveals that deep learning techniques and fiber Bragg gratings are extensively researched in infrastructure, particularly for structural health monitoring. However, there is a lack of research on novel materials like graphite for fiber optic sensor design. The study highlights the growing interest in ML applications for fiber optic sensors, with significant advancements in healthcare, structural monitoring, and environmental sensing. Key findings include the use of ML for detecting structural damage, predicting pipeline degradation, and improving sensor accuracy. The research also identifies emerging trends such as the integration of ML with fiber optic sensors for real-time monitoring, fault detection, and predictive maintenance. The study emphasizes the importance of interdisciplinary collaboration and the need for further research on novel materials and advanced ML techniques to enhance the performance and applicability of fiber optic sensors. The research agenda includes exploring new materials, improving sensor design, and developing more efficient ML algorithms for fiber optic sensing applications. The study concludes that the integration of ML with fiber optic sensors holds great potential for advancing various fields, including healthcare, civil engineering, and industrial monitoring.