String matching is a fundamental problem in computer science and has been extensively studied. Searching for all occurrences of a pattern in a text is a fundamental problem in many applications, like natural language processing, information retrieval, pattern recognition and computational biology. Many string matching algorithms are existing and work efficiently with different applications in different life scopes; one of these algorithms is the Intelligent Predictive String Search Algorithm, this algorithm searches through a given text to find the first occurrence of a pattern without a pre-processing phase that included in many string marching algorithms to calculate the pattern shift values which lead less computations and uses simple rules during a match or mismatch of a pattern character using one sliding window.In this paper we updated the Intelligent Predictive String Search Algorithm three times resulting with three versions; in the first one we reversed the search direction to be from right using one sliding window while in second version we use two sliding windows to scans the text from both sides sequentially and finally we parallelize this version using real parallel environment. Besides, it is easy to parallelize the new developed algorithm gain significant enhancement in decreasing time and memory requirements.

String matching problem is one of the most essential problems in many computer science fields, such as DNA analysis, artificial intelligence, internet search engines and information retrieval. Today, the speed and performance of string matching algorithms is critical and must be improved to meet recent developments in hardware processing environments. The improvement in performance gained by the use of a multi core processor depends very much on the software algorithms used and their implementation. However, the most important factor when writing a parallel algorithm is the fraction of the algorithm that can run simultaneously on multiple cores. In this paper, an efficient algorithm for string matching, Enhanced Pattern Matching Algorithm with Two Sliding Windows (ETWS), is adapted to be implemented under a real parallel environment (PETWS), to enhance the performance of the sequential algorithm through providing less execution time to make it more suitable for today's applications.