In database management systems, the estimation of selectivity is required to predict the cost of possible alternative execution plans for a given query. Many studies show that even very small inaccuracies in the cost estimates can cause the query optimizer to choose a suboptimal execution plan. With the explosion of the internet and text-based data, there is an ever greater need to evaluate queries involving string matching. The main difficulty in handling text data is due to the presence of unclean data, different spellings, and typographical errors. Consequently, there is an increasing demand of more support for good selectivity estimation techniques for approximate string queries. This project aims to develop new approaches for the selectivity estimation of approximate strings using data mining methods.