随着现代信息和网络技术的快速发展，采集的多源异构数据规模迅速增大，有效管理和使用这些数据需要对其进行组织优化和预处理。索引是重要的预处理方法，广泛用于数据的处理和存储管理等领域。以后缀数组为核心的索引统称为SA索引，是全文搜索的重要数据结构。项目研发高效的大规模动态后缀索引算法，成果具有先进的理论和实际性能，可用于海量异构动态数据的在线全文搜索。围绕以下若干关键算法问题开展研究：1、共享内存SA构建并行算法；2、SSD的SA构建外存算法；3、线性时间的SA合并算法；4、同时构建和验证SA的算法；5、自适应的SA构建和合并算法；6、外存SA的模糊查找算法。预期成果：发表SCI/EI收录国际期刊论文6篇（CCF A类期刊论文3篇以上）和会议论文4篇，申请发明专利6个，登记软件著作版权4项；通用于信号、日志、代码和基因等异构动态数据的全文搜索原型系统。

Driven by the rapid development of modern information and network technologies, the amount of multi-source heterogeneous data is growing quickly, optimizing organization and pre-processing of the data are needed for the efficient management and use of data. Indexing is an important pre-processing method that is widely applied in the fields of data processing and storage management. An index built on a suffix array is usually termed as an SA index, it is a fundamental data structure for full-text search. This project researches efficient algorithms for massive dynamic suffix indices, these new algorithms are of advanced theoretic and practical performance and can be applied for online full-text search on massive heterogeneous data. In particular, the key algorithmic problems to be studied are: 1) Parallel algorithms for building an SA in shared internal memory; 2) Algorithms for building an SA in SSD external memory; 3) Algorithms for merging SAs in linear time; 4) Algorithms for building and checking an SA simultaneously; 5) Adaptive algorithms for building and merging SAs; 6) Algorithms for approximate search on SAs in external memory. The expected outcome includes: 6 international journal papers (not less than 3 papers on journals listed as CCF A) and 4 conference papers indexed by SCI/EI , 6 invention patent applications, 4 software copyright registrations; a full-text search prototype system for heterogeneous dynamic data of signals, logs, codes and genome.