1) The ability to recognise the individual voice patterns and sound waves and to interpret this information to complete a meaning action.
2) By means of training the software to recognise particular sound waves and informning that software of the required action on recognising this sound wave.
3) key uses i would see are for disabled persons who are unable to use a mouse or keyboard, this will give them the power they require to complete these tasks. Also for a basic form of security, you could require voice recognition to enable a login to a network or system.
4) One package that is available is Dragon Naturally Speaking, this is a wide range of software depending on your overall needs and budget, and by means of hardware, if you by a cheapy sound card and a poor quality microphone, you will expereince extended difficulty in training the software. Where as higher qaulity sound card and microphone will allow for more consistent input to the system.