Fuzzy Association Rule Classification System For Human DNA profile Identification

Abstract

Recently, numerous identification methods are adopted in human identification systems particularly for forensic affairs. These methods are used to reduce the human efforts, related errors, cost, and consumed time in addition to increase the accuracy of human identification systems. The definition of STR DNA profile of a person’s, is associated to the reduction or detection of criminal investigation associated to the identification of the suspect, or to an investigation in counter terrorism as well as in the National Security Service. DNA is used for disaster victims or suspected individuals as a main means of identification. It possesses a high degree of verification in the courts when there is biological evidence, and is also used to identify persons accused and innocent people in criminal investigation with incredible accuracy. This thesis proposes a forensic human identification system based on Deoxyribonucleic Acid (DNA) classification. This system is depend on the concept of fuzzy logic and data mining in identification, that play role important in established a plenty of effective applications in DNA profile analysis and human identification, and reduce error or prevent and so they are chosen for their high accuracy. The main contribution of the presented system is, used association rules for classification the DNA motifs in addition to the fuzzy inference to correct errors, that can offer a high range of useful information. This information is permanently used to solve the crimes and identify damaged bodies. Additionally, this overcomes the lack problem in database information. It is important to note that the proposed system includes DNA based human identifications to cover all investigation cases. The aim of this thesis is to build a forensic human identification system that uses data mining techniques to access dataset by enforce particular algorithms for discovering advisable information or good patterns from present datasets for certain objective. To improve the data mining work, fuzzy logic is used to increase the flexibility for supporting DNA profile forensic human identification in making decisions. It utilizes sixteen Short Tandem Repeats (STRs) DNA loci used in identification methods for Population. It is important to note that the proposed system has been verified In terms of matching of provided samples with the samples stored in the database through conducting several experiments. The results were satisfactory and characterized by large percentage and high accuracy. Finally, performance of this system is evaluated and in turns the proposed system proves its capability in forensic human identification and scalability to handle huge amounts of data.