Browse Prior Art Database

Automation of point of purchase processing using multimodal interaction

IP.com Disclosure Number: IPCOM000124639D
Original Publication Date: 2005-May-03
Included in the Prior Art Database: 2005-May-03
Document File: 4 page(s) / 103K

Publishing Venue

IBM

Abstract

This disclosure describes the addition of a multimodal component to an automated point of purchase processing system. The addition of a multimodal interaction component enhances the user's experience of point of purchase processing by making entering purchase information into the user's device easier, more accurate and more reliable.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 29% of the total text.

Page 1 of 4

Automation of point of purchase processing using multimodal interaction

Multimodal interaction is most useful for small hand-held devices such as PDA's and cellphones. On these devices it is not easy to enter text information using a keypad and stylus. For cellphones entering text is even more difficult because there is no stylus and the keypad is dedicated primarily to entering telephone numbers. Multimodal interaction further enhances the user's experience of point of purchase processing in three ways. First, the service provider provides a multimodal interface to its point of purchase application displayed on the device. This interface includes speech input grammars that allow the user to speak the information requested by the provider. Because the interface is multimodal, the information that the user should speak is visually presented. Text-to-speech prompting also reinforces what the user should say. Second, multimodal auto fill allows the user to easily enter profile information such as name, address, telephone number and account information. Multimodal auto fill will work on devices where the speech recognition and text-to-speech is on the network as well as located on the device. Finally, point of purchase processing requires client authorization and authentication to be secure. The multimodal component adds speaker verification and authentication as an efficient means to authenticate the user using a multimodal interface.

The figure below describes an automated point of purchase system with an additional component for providing multimodal interaction with the application on the client device. The component consists of: 1) at least one more server for providing the multimodal auto fill grammars and voice print data to the application, 2) speech engines for delivering text-to-speech output and recognizing speech input, and 3) an optional voice browser for interpreting the application XML markup. There is also a secure database of client voice prints and an engine that matches the user's speech input with a voice print. For the figure below, the service provider server serves a multimodal interface to the client's device. Here the speech recognition and text-to- speech engines are also located on the service provider server. However, the speech recognition and text-to-speech engines can be distributed or located on the user's device.

One example of a point of purchase system is a restaurant that takes reservations and orders over the internet. This example illustrates how a multimodal component enhances the user experience of interacting with an automated point of purchase system. For this restaurant scenario the register/order/pay scenario can be described in terms of a sequence of multimodal interactions. The customer first registers his or her name with the receptionist by saying a phrase such as "my name" to the device. Saying "my name" selects the customer's name from the user's profile information, and it is sent to t...