Indeed automated speech recognition (ASR) is still an evolving technology. The nightmare scenario for ASR on a mobile phone is in speaker mode. The moment #KINJO delivers its welcome message, it gets in listening mode to receive the verbal command from the user. In speaker mode, the system hears itself, doesn't understand the command and takes time to process it while in the meantime the user first gently and then forcefully from frustration, keeps instructing the system. It then all collapses. If ever you got frustrated that way, it is understandable but simply don't do it. It seems a reasonable limitation.

Background noise is also an impediment to ASR but the new bluetooth headsets are quite good at filtering some noise out.

You should try again and let us know. Nevertheless, AskKinjo will keep upgrading its ASR with ever better software.

As for the profile, it is true that part of the requested information is to deliver targeted advertising to the user. Since advertising is needed to support the #KINJO free service, it is better to deliver an ad most relevant to the users based on their profile. The users thus get a free service and an ad to which they can relate.

Like or Dislike: 0  0