Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2001-02-26
2003-09-23
Rones, Charles (Department: 2175)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000
Reexamination Certificate
active
06625613
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to dialogue processing systems, and more particularly to frame representations, SQL queries, and relational databases for dialog processing systems.
BACKGROUND
Natural language dialogs will play an increasingly important role as an interface to computer systems in general, and especially as an interface to the internet. Computer systems that will rely increasingly on natural dialog systems include automated call handlers, ticket machines, automatic teller machines (ATMs), reservation systems, and especially interactive on-line internet services. In fact, natural language dialogs will play an increasingly important role in any other system involving human-computer interaction which can be characterized as a dialogue. However, one of the barriers to the wide adoption of this technology is the lack of a domain-portable dialog system.
In natural language dialog systems it is common to represent domain knowledge in terms of frames, where a frame includes a set of feature-value pairs describing a domain object. Since a frame has a simple list of feature-value pairs, it does not contain any information regarding the interdependence of its feature values (feature values are interdependent because specifying the values of one or more features restricts the range of possible values for the other features). Given a frame representation, it is desirable that a dialog system be able to infer the possible values of features based upon the values of other features. However, it has been difficult to devise a technology capable of inferring values of features that is not domain specific.
To the extent that existing natural language dialog systems perform domain reasoning, they generally employ domain-specific heuristics, which involves hand-crafting a set of rules and procedures for each application domain. Simply stated, the problem is how to construct a computationally efficient domain-portable mechanism for using domain knowledge to specify the values of certain features of a frame representation based upon the values of other features in the same frame.
In contrast to previous attempts at dialog management systems which generally rely on domain specific heuristics, the current method is automatic and domain-independent, relying only upon the structure of the frame representation and associated rational database. It is also simple and efficient, relying upon standard and widely-available relational database technology.
REFERENCES:
patent: 5201046 (1993-04-01), Goldberg et al.
patent: 6016497 (2000-01-01), Suver
patent: 6035275 (2000-03-01), Brode et al.
patent: 6044347 (2000-03-01), Abella et al.
patent: 6411951 (2002-06-01), Galindo-Legaria et al.
patent: 6470333 (2002-10-01), Baclawski
patent: 2002/0062312 (2002-05-01), Gupta et al.
patent: 2002/0083039 (2002-06-01), Ferrari et al.
M. Denecke, Interactive Systems, Inc. and A. Waibel, Interactive Systems, Labs.; “Integrating Knowledge Sources for the Specification of a Task-Oriented Dialogue System”; (Pittsburgh, Pa; pp. 33-39) (no date).
D.G. Bobrow, R.M. Kaplan, M. Kay, D.A. Norman, H. Thompson and T. Winograd; “Gus, A Frame-Driven Dialog System”;Artificial Intelligence8 (1977), pp. 155-173.
Lamb James A.
Motorola Inc.
Rones Charles
LandOfFree
Automatic generation of SQL for frame completion does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Automatic generation of SQL for frame completion, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automatic generation of SQL for frame completion will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3043595