Continuous language-based prediction and troubleshooting tool

Details Continuous language-based prediction and troubleshooting tool Continuous language-based prediction and troubleshooting tool

2001-04-30

2004-08-03

Baderman, Scott (Department: 2184)

Error detection/correction and fault detection/recovery

Data processing system error or fault handling

Reliability and availability

C714S057000, C705S026640, C707S793000

Reexamination Certificate

active

06772374

ABSTRACT:

CROSS-REFERENCE TO RELATED APPLICATIONS
Not Applicable.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not Applicable.
REFERENCE TO AN APPENDIX
Not Applicable.
BACKGROUND OF THE INVENTION
(1) Field of the Invention
The present invention relates generally to technical support of commercial and installed-base products, and more specifically to an automated predicting and troubleshooting tool for installed-base apparatus.
(2) Description of Related Art
Currently, it is highly expensive for equipment manufacturers, particularly computer apparatus original equipment manufacturers (“OEM”) and their related companies, to support installed-base equipment. Typically, a customer, for example, an end-user of a Hewlett-Packard® ink-jet printer, telephones a call agent to report a problem. This call agent guides the customer through a troubleshooting sequence—generally in a question-and-answer format—that hopefully leads to resolution of the problem or at least an identification of the likely cause. The question-and-answer sequence is a knowledge-based system based on prior experiences which have been formatted into case base, decision tree, troubleshooting flow charts, and the like as are commonly known, for use by the manufacturer's call agent. Furthermore, the session is likely to be based on the individual expertise of the call agent handling the immediate matter.
Attempts are on-going to try to automate such systems. One such attempt involves the use of Bayesian networks (a directed acyclic graphical representation of causal relationships between variables, associating conditional probability distributions to variables given their parents). Pending patent applications assigned to the common assignee herein, Hewlett-Packard Company, teach A METHOD FOR KNOWLEDGE ACQUISITION FOR DIAGNOSTIC BAYESIAN NETWORKS (Skaaning et al., Ser. No. 09/261,769, filed Mar. 3, 1999) and AUTOMATED DIAGNOSIS OF PRINTER SYSTEMS USING BAYESIAN NETWORKS (Ser. No. 09/353727, filed Jul. 14, 1999, by Skaaning et al.).
Existing software troubleshooter agents take input from a question-and-answer dialogue as the series of provided answers, typically in multiple choice format. The order of the questions is controlled by the software, treating the end-user like a “subroutine” to obtain diagnostic information. This can be frustrating and inefficient, particularly in situations where the end-user has a good idea of what the actual fault may be. Few, if any, existing automated troubleshooter agents or programs take unstructured text for diagnostic input purposes.
Thus, there is a need for an automated system and methodology working in real-time to troubleshoot problems and predict probable solutions.
BRIEF SUMMARY OF THE INVENTION
The present invention provides a free-form text or voice input system for troubleshooting and problem resolution that restores control to the end-user. The need is fulfilled in an automated free-form data input and analysis format without ruling out suggestive question-and-answer aid. The present invention can provide for a mixed initiative troubleshooting session, where both human and computerized interface can indicate next logical steps of the troubleshooting and potential resolution prediction process.
More specifically, the present invention provides a prediction/troubleshooting tool, working progressively in a real-time based systematic manner on “clues” or “hints” appearing in a stream of plain text (or speech if voice-recognition is employed).
In its basic aspect, the present invention provides a computerized diagnostic system including: a data base correlating product information with keywords related to failure modes; and an interface program for receiving plain text related to the failure modes and for comparing said plain text to the keywords and for predicting probabilities of potential solutions to said failure modes based upon said comparing.
In another aspect, the present invention provides a method for troubleshooting product failures, the method including: providing a data base correlating product information with keywords related to failure modes; using a network link, reporting at least one specific one of said failure modes of at least one product; in real time, comparing correlated product information from said data base to said reporting at least one specific one of said failure modes, and based on said comparing, predicting probabilities of effectiveness of potential solutions to curing said at least one specific one of said failure modes.
In still another aspect, the present invention provides a computer memory including: program code associated with a network server and client system, said program code including subroutines associated with said client providing a plain text report of a failed client subsystem, subroutines associated with said server for comparing said plain text report to a predetermined data base having individual units of the subsystem correlated with keywords associated with failure mode descriptions, and subroutines for using said comparing and generating a report to the client of probabilities of selected solutions to the failed client subsystem.
In yet another aspect, the present invention provides a method of doing business with respect to troubleshooting and providing solutions to reported failures of installed base products, wherein an entity doing said business maintains a data base of product information correlated to failure mode scenario information as represented by keywords, the method including: establishing a server having an internet access site;
providing a resolution prediction troubleshooting tool at said site; allowing client access to said site via a browser; following client access of the site, receiving a client generated report of at least one specific one of said failures of at least one installed base product; and in real time, comparing the keywords to said report, based on said comparing, predicting probabilities of potential solutions to curing a specific failure represented by said report, and transmitting said probabilities to said client.
Note that the invention allows combining with business information and data, such as costs and inventory, to recommend specific action(s).
The foregoing summary is not intended to be an inclusive list of all the aspects, objects, advantages, and features of the present invention nor should any limitation on the scope of the invention be implied therefrom. This Summary is provided in accordance with the mandate of 37 C.F.R. 1.73 and M.P.E.P. 608.01(d) merely to apprise the public, and more especially those interested in the particular art to which the invention relates, of the nature of the invention in order to be of assistance in aiding ready understanding of the patent in future searches. Objects, features and advantages of the present invention will become apparent upon consideration of the following explanation and the accompanying drawings, in which like reference designations represent like features throughout the drawings.


REFERENCES:
patent: 5826008 (1998-10-01), Bluvband
patent: 6212494 (2001-04-01), Boguraev
patent: 6230287 (2001-05-01), Pinard et al.
patent: 6389561 (2002-05-01), Bluvband
patent: 6442542 (2002-08-01), Ramani et al.
patent: 6487677 (2002-11-01), Jantz et al.
patent: 6513129 (2003-01-01), Tentij et al.
patent: 6532426 (2003-03-01), Hooks et al.
patent: 6633782 (2003-10-01), Schleiss et al.

Affiliated with

Also associated with

Rating

Continuous language-based prediction and troubleshooting tool does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Continuous language-based prediction and troubleshooting tool, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Continuous language-based prediction and troubleshooting tool will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3328230