Parallelizing the Sparse Matrix Transposition: Reducing the Programmer Effort Using Transactional Memory | Publicación