System and method for labeling alert messages from devices for automated management

• US 9,569,327 B2
• Filed: 10/03/2012
• Issued: 02/14/2017
• Est. Priority Date: 10/03/2012
• Status: Active Grant

• Current Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Singh, Anand, Hoppenot, Yves, Roulland, Frederic, Valobra, Pascal, Ciriza, Victor
• Primary Examiner(s): SPOONER, LAMONT M
• Application Number:
  US13/644,187
• Publication Number:
  US 20140095144A1
• Time in Patent Office: 1,595 Days
• Field of Search: 704 1- 10
• US Class Current: 1/1
• CPC Class Codes: G05B 23/027 Alarm generation, e.g. communication protocol; Forms of alarm
• CPC Class Codes: G06F 11/3013 where the computing system is an embedded system, i.e. a combination of hardware and software dedicated to perform a certain function in mobile devices, printers, automotive or aircraft systems testing or monitoring of control systems or parts thereof G05B23/02
• CPC Class Codes: G06F 11/3079 the data filtering being achieved by reporting only the changes of the monitored data
• CPC Class Codes: G06F 11/3086 where the reporting involves the use of self describing data formats, i.e. metadata, markup languages, human readable formats
• CPC Class Codes: G06F 3/1203 Improving or facilitating administration, e.g. print management
• CPC Class Codes: G06F 3/121 Facilitating exception or error detection and recovery, e.g. fault, media or consumables depleted
• CPC Class Codes: G06F 3/1234 Errors handling and recovery, e.g. reprinting G06F3/1261 takes precedence
• CPC Class Codes: G06F 3/1285 Remote printer device, e.g. being remote from client or server
• CPC Class Codes: G06F 40/56 Natural language generation
• CPC Class Codes: H04L 12/1895 for short real-time information, e.g. alarms, notifications, alerts, updates

Abstract

An alert processing system and method are adapted for processing device alerts. The system includes a routing device in communication with a printer. The routing device receives at least one alert description in a source language transmitted from the printer. The routing device identifies a set of words derived from the alert description related to a condition of the associated device. The routing device compares the set of words, in a target language, to a categorization model and, based on the comparison, categorizes the set of words into to one of a predetermined set of alert categories.

Claims

1. An alert processing system comprising:

a communication interlace adapted to receive at least one alert description in a natural language of an alert message output by an associated device, the alert description defining a problem with the associated device;

an alert filtering module adapted to;

access a stored referenced alerts list to determine if the alert description corresponds to one of a set of previously referenced alert descriptions,

when the alert description is determined to correspond to a previously referenced alert description, labeling the alert description with the label of the corresponding previously referenced alert description, and

when the alert description is determined not to correspond to any of the previously referenced alert descriptions, labeling the alert description as an unreferenced alert description;

a word exclusion module adapted to extract a set of words from the alert description related to a condition of the associated device;

an alert categorization module adapted to categorize the alert description, when it has been labeled as unreferenced by the alert filtering module, into to one of a predetermined set of alert categories of a categorization model based on the extracted set of words;

a language module adapted to;

compare each word in the set of words with a plurality of stored dictionaries in different languages for identifying at least one source language of the word,

assign a source language to the set of words based on the identified source languages of the words, and

translate the set of words from the assigned source language to a target language; and

a processor adapted to implement the modules.

2. The alert processing system of claim 1, wherein the condition indicates a status of at least one component of the associated device.

3. The alert processing system of claim 1, wherein the associated device comprises a printer.

4. The alert processing system of claim 1, wherein the word exclusion module includes instructions for removing selected content from the alert description, the selected content being selected from the group consisting of:

error codes;

special characters;

appended content;

articles;

conjunctions;

adpositions; and

combinations thereof.

5. The alert processing system of claim 1, wherein the categorization module is trained to categorize sets of words in the target language.

6. The alert processing system of claim 1, further comprising a rule module adapted to:

determine a rule corresponding to the alert description based on the categorization.

7. The alert processing system of claim 1, further comprising an update module adapted to:

provide the categorized alert description to an advisor for review;

acquire one of a validation and correction provided by the advisor; and

based on the one of the acquired validation and correction, add the categorized alert description to a referenced alerts list.

8. The system of claim 7, wherein further comprising a training module for retraining the categorization module at intervals based on the categorized alert descriptions added to the referenced alerts list.

9. The alert processing system of claim 1, wherein the system receives alerts generated by a set of devices, the associated device being one of the set.

10. A method for processing alerts transmitted by at least one device, the method comprising:

training a categorization model, comprising;

receiving sample alert descriptions as text, the sample alert descriptions being representative of types of new alerts to be processed by the trained categorization model, and receiving a manually-assigned category in a predefined taxonomy for each sample alert description,

extracting a set of words from each sample alert description, and

training the categorization model with the set of words and the manually-assigned category of each of the sample alert descriptions to assign each word in a remaining set of words of a new alert description to a given category in the taxonomy or probabilistically over all categories and based on the word assignments, to assign an overall category to the new alert description;

receiving at least one new alert, transmitted by the at least one device, the new alert using predefined codes and text to define a problem with the associated device outputting the new alert, the new alert including a new alert description in a natural language;

extracting a set of words from the new alert description related to a condition of the associated device;

with a processor, categorizing the new alert description into to one of the predetermined set of alert categories with the trained categorization model, based on the extracted set of words from the new alert description, and

outputting the categorized alert description.

11. The method of claim 10, further comprising:

accessing a stored referenced alerts list to determine if the alert description corresponds to one of a set of previously referenced alert descriptions;

for an alert description determined as not corresponding to any of the previously referenced alert descriptions, labeling the alert description as an unreferenced alert description to be processed by the categorization module; and

for an alert description determined to correspond to a previously referenced alert description, labeling the alert description with the label of the corresponding previously referenced alert description such that the alert description is not processed by the categorization module.

12. The method of claim 10, further including:

assigning a label to the new alert description based on the alert category; and

identifying a rule for handling the alert based on the label.

13. The method of claim 12, further including responding to the new alert description based on the rule.

14. The method of claim 10, further comprising identifying a source language for the new alert description and translating the alert description form the source language to a target language in which the categorization is performed.

15. The method of claim 14, wherein the identifying the source language comprises:

comparing each word in the set of words with each of a plurality of dictionaries, each dictionary being in a respective language, to determine if the word is present in the dictionary;

computing a probable source language based on the comparing; and

assigning the probable source language to the set of words.

16. An alert processing system for handing device alerts, the system comprising:

memory which stores instructions for performing the method of claim 10 and a processor, in communication with the memory for executing the instructions.

17. A computer program product comprising a non-transitory recording medium which stores instructions for performing the method of claim 10.

18. The method of claim 10, wherein the extracting a set of words from each sample alert description includes filtering out computer-readable codes.

19. A system for processing device alerts transmitted from an associated device, the system comprising:

memory storing;

a plurality of dictionaries each corresponding to a respective one of a set of source languages, and

a language module for identifying a source language of a new alert description by comparing each of a set of words extracted from the new alert description to each of the dictionaries and, where the source language is not a target language, translating the set of words from the source language to the target language;

a categorization module trained on sets of words from labeled prior alert descriptions, the words being in a target language, the categorization module assigning a category to the new alert description based on the set of words extracted from the new alert description in the target language; and

an update module which updates a referenced alerts list based on optionally manually validated assigned categories of a plurality of alert descriptions processed by the categorization module, the plurality of alert descriptions including the new alert description, the referenced alerts list being provided to an associated filtering module for assigning categories to uncategorized alert descriptions without using the categorization module; and

a processor in communication with the memory for implementing the modules.

20. The system of claim 19, wherein the memory further stores:

a set of rules for responding to the new alert description; and

a rule module for identifying one of the set of rules for handling the alert based on the categorization.