USE OF VIOCE RECOGNITION SYSTEMS TO MAKE RESTURANTS WORKING FASTER

By Sajida

As we go through the very informative seminar on the MICROS POS .that showed us very effective information system to make business process quite effecient.with respect to that i have searched to make the resturant systems more better i have found one the use of vioce recognition system which has already been applied in some resturants and this software named as SIRI.Kindly go through it and discuss regarding its applications & implications  in Pakistan if applied in resturants.

 Siri (software):

What can Siri do that other speech recognition technologies cannot?

Speech recognition, the ability of a device to break down a stream of audio into text, is a mechanical process.  Siri does very well in discerning continuous speech. However, It certainly is not currently the best system yet invented. 

The mechanical process of Speech recognition is really not the big story about Siri.  Siri's unique abilities are based on a far more complex issue, understanding what is being asked, the context it is being asked and providing a result.  There are 3 primary foundations that form the infrastructure of Siri:

   * A conversational interface: With the best of speech and semantic language understanding combined with an interactive dialog that helps guide people toward saying what they want to do. The conversational interface allows for much more interactivity that 'phrase" based search style interfaces.  Siri focuses on a better understanding of intent.  If Siri didn’t quite hear what you said, or isn’t sure what you meant, it can ask for clarifying information.  Siri can prompt on ambiguity: did you mean pizza restaurants in Chicago or Chicago-style pizza places near you? Reasonable guesses can also be made based on context. For example if it is lunchtime and the speech interpretation comes back garbled about food Siri may assume you probably meant “places to eat near my current location”. If this assumption isn’t correct, it is easy to correct in a conversation.

    * Semantic auto-complete – This is similar to  â€œautocomplete” search boxes with but Siri integrates a semantic and linguistic model, “semantic completion”.  This makes possible a rather fast result with complex requests, eg. Sushi restaurants in (in the local area) that are open for dinner.   The results are a magnitude better when Siri gets to understand what you use the system for most often.

    * Service delegation – User’s intent is the basis of how Siri works.  It dispatching to multiple heterogeneous services parts of the request and  gathers and integrates the results, This is presenting back to the user as a set of proposed solutions with user prior biases as an answer to their request. In the case of a restaurant, Siri will combine information from many a multitude of sources like local business directories, geospatial databases, restaurant guides (like Yelp), restaurant review sources, online reservation services (like Open Table), and combine this with the user’s own favorite restaurants

Thus the Speech recognition, although very important is just about 1/8th of the entire process in the end.  I wrote in detail about Siri here: Apple Products and Services: Why is Siri important?. 

To illustrate from this post how Siri is different it is important to understand most systems in use are really Keyword based solutions that operate on a very limited Task completion criteria.  An example would be the typical car navigation systems that allows for you to choose from a menu of functions like what are the closest Gas stations.  This is of course very functional, it is not really actively conversational nor does it understand any form of context.

Siri base points are: 

    * Does Things For You- Task completion:

- Multiple Criteria Vertical and Horizontal searches

- On the fly combining of multiple information sources

- Real time editing of information based on dynamic criteria

- Integrated endpoints, like ticket purchases, etc.

    * Gets What You Say- Conversational intent:

- Location context

- Time context

- Task context

- Dialog context

    * Gets To Know You- Learns and acts on personal information:

- Who are your friends

- Where do you live

- What is your age

- What do you like

The existing Speech Recognition systems will continue to exist and will no doubt be productive. The best comparison for existing is that existing Speech Recognition systems are like a Calculator as related in comparison to Siri as an advanced computer.