Methods of Information Extraction and Analysis in Data Stream

With the development of communication technology, information extraction and analysis in data stream becomes more and more universal in embedded pervasive computing systems. The increasing size of the data stream often makes the process of extraction and analysis becomes a bottleneck of the system, especially for embedded system with limited resource. This paper researches on various methods of information extraction and analysis in data stream, analyzes their advantages, disadvantages, applying occasions and ways to optimize the technology. Then it elaborates on optimized algorithms based on load balancing multithreading, which can also serve as a useful reference for the optimization of the key process or thread that may become the bottleneck in the embedded system. The optimization is applied into PTV (personalized TV), a digital set-top box, for the extraction and analysis on PSI and SI information in TS stream. The testing result proves that it can significantly reduce the executing time. At last, it puts forward the improving direction of the optimized algorithms using load-balancing multithreading based on priority.