The Problem of Informational Object Identification in Case of the Considerable Quantity of Identifying Features
| Autor: | S. D. Kulik, D. A. Nikonets |
|---|---|
| Jazyk: | English<br />Russian |
| Rok vydání: | 2010 |
| Předmět: | |
| Zdroj: | Безопасность информационных технологий, Vol 17, Iss 1, Pp 86-87 (2010) |
| Druh dokumentu: | article |
| ISSN: | 2074-7128 2074-7136 |
| Popis: | The modification of the algorithm of identification of the informational object, used for identification of the hand-written texts performer in an automated workplace of the forensic expert, is presented. As modification, it is offered to use a method of association rules discovery for definition of statistically dependent sets of feature of hand-written capital letters of the Russian language. The algorithm is approved on set of 691 samples of hand-written documents for which about 2000 identifying feature are defined. The modification of the identification algorithm allows to lower level of errors and to raise quality of accepted decisions for information security. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|