Audio Files Present Challenges For Computer Forensics and E-Discovery
Unified Communications is the term for the integration of all communications services - voice and data communications - through the Internet. You can also data from its myriad forms such as e-mails, instant messaging, data from the economy applications, faxes and SMS messages. But among the main sources of voice network or on digital media such as VoIP (Voice over Internet Protocol), Voice-mail, audio, video, Web conferencing, white, and. Wav files. Such an integrated communication can save money on the exploitation of households.
savings result from, among other charges, the abolition of taxes for long-distance calls when using VOIP, the waiver of the need for a trip to meetings, but if they are under a virtual environment, or distant travel in the classroom if a teacher or a team can White Board physically different sites. Savings, as are 26% of companies they have adopted. But if claims disputes recognizable data. Wav files and voice can be difficult and expensive for a computer Forensic expert or an e-discovery system for searching and indexing.
There are many tools for the detection of text files, and even for the text deleted files. These range from computer forensics and packaged as a result of access Forensic Toolkit, that cost thousands of dollars for Open Source tools, including publishers hexa, l & # 39; user that cost nothing at all. The more important lots may be less costly in the long run, if people are billable in the mix.
There are a lot of e-wild dear to the discovery of support systems for the storage and indexing the vast mass of data, are on a daily basis in the business. The services can be outsourced or made in companies. In addition, the costs of the implementation of systems and procedures on the ground May pale against sanctions and fines could follow, is not ready for litigation, should they occur.
There are also many effective tools for digitizing paper documents into text files, which will be analysed.
While many tools for research and data storage effective, and it is precisely when it comes to audio, the absence of such a level of precision or simply, there are still for the specific purpose of research. Currently, there are three avenues of research audio: phonetic research, processing of the hand, and automatic transcription.
Phonetics matches Search technology wave, or phonemes, a library of known models airwaves. For example, the abbreviation "B2B" was represented by the phonemes: "_B _IY _T _UW _B _IY" (Wikipedia Nexidia example, a company involved voice recognition). Given the vast differences in patterns of taking speech, pronunciation, dialects and accents, the precision of this method is spotted. It produces many false results. And while it can identify sections and expressions of interest, it is not to transcribe the audio-text - the audio signal must be owned.
Manual transmission audio recordings so that transcription of the text may then be automatically sought is lengthy. As it depends, with an audience of the construction of words, when they heard this intense task can also be very costly. There may be security, it is the audio outside the company (or maybe the country).
Machine transcription is a means of automatic conversion of the audio text. But he suffers precision. It compares "is part audio libraries, with discussion of diversity issues, not with regard to libraries and clarity of the shooting. While high-quality recordings can be adapted for the recognition of 85% or (a number of positive points in comparison with nearly 100% accuracy of finding pure text), when handling with voicemail, accuracy dips down as low as 40%.
The new regulation of the Federal Civil Procedure (FRCP), require that companies have a means of identification of the main communication and data sources. These data must then be recorded. For reasons of efficiency, both in optimizing the height of storage necessary, and the decrease in the volume of data that must be identified and, for litigation, it is also important that we Specifically, to identify the data is useless.
While the requirements for data retention and increase storage costs, to discover what his treatment, and what should be deleted can be costly. Given that digital information must be stored and indexed (or sought does). That technology is not yet mature and grows. Perhaps there is an opening for an innovative company to prosper here, especially if the situation, some kind of breakthrough in the voice to text technology. In the meantime, businesses face a difficult question to decide what goes and what remains.
