Genetik algoritma parametrelerini kullanarak SegGen tematik segmentasyon algoritmasının geliştirilmesı

Due to the remarkable increase in the number of available text databases in last decades, the need for efficient searching methods has become a major challenge for information retrieval. Thematic segmentation can be defined as the process of separating written texts into meaningful homogeneous units according to the criteria stated in Salton?s definition which states that thematic segmentation of a text is its splitting into segments such that the internal cohesion of segments and the dissimilarity between adjacent segments are maximum. SegGen is a linear thematic segmentation algorithm grounded on a variant of the SPEA and aims at optimizing the two criteria of the Salton's definition of segments: a segment is a part of text whose internal cohesion and dissimilarity with its adjacent segments are maximal. This thesis describes improvements that have been implemented in the approach taken by SegGen by tuning the genetic algorithm parameters according with the evolution of the quality of the generated populations. Two kinds of reasons originate the tuning of the parameters. The first one rests on autonomous search, which consists in modifying the parameters and operators of the genetic algorithm along with the increasing quality of the generated population through the generations. The second one is also to consider the increasing of quality of the population as the process evolves, but to do so considering that the nature of the coding of individuals which in this case are segmentation instances represented by binary vectors corresponding to the positions of the boundaries of the segmentations.