CUSMART: effective parallelization of string matching algorithms using GPGPU accelerators

• Introduces CUSMART, a CUDA-based parallel implementation of the SMART library, integrating 64 string matching algorithms for GPGPU acceleration. • Demonstrates significant performance improvements and scalability of string matching algorithms on NVIDIA GPUs through systematic parallelization and optimization. • Provides a comprehensive comparative analysis of parallelized algorithms across multiple scenarios, identifying strengths and weaknesses for specific use cases. • Establishes practical optimization techniques for GPGPU-based string matching, contributing to high-performance text processing in big data and related fields.