Abstract:
|
The suffix tree is the most powerful and versatile structure in the
string matching. For this structure, the classical construction
algorithms in linear time are the Weiner, the McCreight and the
Ukkonen algorithm. In this work is presented a review of these
construction algorithms and a new one that improves two non functional
characteristics: efficiency and usability . This new algorithm reduces
the construction time in about 50% (experimentally tested) , the query time
and space spent allowing to work with much
longer sequences. In this way, the algorithm uses a new design of
suffix links structure that provides a major locality between the
subtrees (a profit example will be shown). Further more, the
resultant algorithm are much more understandable, making it
specially easy to modify for resolve concrete problems. |