The Use of Speech Technology to Protect the Document Turnover

Abstract

The wide current paper documents implementation in practice workflows are shown. The basic aspects of document protection related to the protection of their content and legal components are underlined. For contextual component assigned semantic information aspect of the document is considered. For legal component attributed facts and conditions for the creation, approval, negotiation of the document to specific persons is viewed. The documents protection problem importance is shown in connection with possible terrorist threats. The importance of such factor as the time of fraud detection towards the efficiency of documents protection is shown. The fraud detection time requirements for documents of different nature – financial, legal, management is analyzed. The documents used for the operational management of dangerous objects is point out as the most sensitive to the falsification. It is shown that their deliberate falsification can lead to accidents and technogenic catastrophes and human casualties. A comparative analysis of currently used protecting documents methods are presented. Biometric and non-biometric methods of documents protection are point out.Theanalysis of their short comings are given. The conclusion about the prospects of document protection on the basis of the voice signature technology are done. The basic steps of voice information processing in the implementation of this technology are analyzed. The software that implements a documents counterfeiting new protection technology is proposed. The technology is based on the audiomarkers usage at the end of the document, which contains a general information about it. The technology is applicable to the wide range of documents such as financial and valuable papers, contracts, etc. One of the most important advantages of this technology is that any changes in the document can not be done without the author of the document because audiomarker keeps the biometric data of the person. The developed software is oriented on the mobile applications on the basis of android operating system. The software has a friendly interface, which allows one to create, correct and check audiomarkers. The performed software tests shows its high efficiency, including the possibility of audiomarkers recognition in various lighting conditions. The possibility of document protection level improvement by logging in voice signature the current level of mental and emotional stress of the document author, which can be used to detect its inadequate state is shown. This makes it possible to identify cases of document preparation under the application of physical or mental violence.

Keywords

• spectrogram
• audiomarker
• recognition
• documentation
• graphic image