c © Copyright by Taewook Oh, 2015.

Parallelism may reside in the input of a program rather than the program itself. A script in-terpreter, for example, is hard to parallelize because its dynamic behavior is unpredictable until an input script is given. Once the interpreter is combined with the script, the resulting program becomes predictable, and even parallelizable if the input script contains paral-lelism. Despite recent progress in automatic parallelization research, however, existing techniques cannot take advantage of the parallelism within program inputs, even when the inputs remain fixed across multiple executions of the program. This dissertation shows that the automatic exploitation of parallelism within fixed pro-gram inputs can be achieved by coupling program specialization with automatic paralleliza-tion techniques. Program specialization marries a program with the values that remain in-variant across the program execution, including fixed inputs, and creates a program that is highly optimized for the invariants. The proposed technique exploits program special-ization as an enabling transformation for automatic parallelization; through specialization, the parallelism within the fixed program inputs can be materialized within the specialize