System and method of information extraction, search and summarization for service action recommendation

Patent metadata

Field	Value
Publication number	US-12579154-B2
Application number	US-202218051889-A
Country	US
Kind code	B2
Filing date	Nov 2, 2022
Priority date	Nov 2, 2022
Publication date	Mar 17, 2026
Grant date	Mar 17, 2026

Abstract

A service intelligence system is disclosed. The service intelligence system receives a search query, identifies anomalous machine codes for a machine associated with the search query, selects one or more historical data elements based on a similarity score between the identified anomalous machine codes and machine codes associated with the one or more historical data elements, ranks the one or more historical data elements based on the similarity score, selects a predetermined number of the one or more historical data elements based on descending rank order, and provides at least one corrective action or at least one part replacement associated with the tag for each of the predetermined number of highest ranked historical data elements in response to the search query. The machine codes are identified for a predetermined time period and each of the one or more historical data elements is associated with a tag.

First claim

What is claimed is:

1 . A service intelligence system, comprising:
a non-transitory memory having instructions stored thereon; and a processor configured to read the instructions to:
receive, over a network, a search query associated with an occurrence of a device error;
search a database to identify a first set of historical data elements each associated with one or more errors similar to the device error, wherein the search includes a natural language search process based on one or more terms extracted from the search query, wherein each historical data element in the first set of historical data elements is associated with one or more first tags;
determine a machine identifier from the search query;
retrieve machine log events of a machine associated with the machine identifier for a predetermined time period that starts before the device error occurred and ends after the device error occurred;
filter the machine logs to identify one or more anomalous log events;
search a database based on the one or more anomalous log events associated with the machine identifier to identify a second set of historical data elements having historical anomalous machine log events similar to the one or more anomalous log events;
determine a similarity score between the device error and the historical service tickets based on a correspondence between the anomalous machine log events and the historical machine log events associated with the second set of historical data elements, wherein each of the one or more historical anomalous machine log events is associated with one or more second tags;
generate a ranking of the second set of historical data elements based on the similarity score;
generate a result set including at least one corrective action or at least one part replacement associated with the one or more first tags and at least one corrective action or at least one part replacement associated with the one or more second tags associated with a highest ranked subset of the second set of historical data elements;
remove contextual stop words from the result set by truncating a heavy tailed distribution of the result set; and
transmit, over the network, the result set in response to the search query.

2 . The service intelligence system of claim 1 , wherein the similarity score is a value between 0 and 1, and wherein the similarity score for a corresponding historical data element of the second set of historical data elements is calculated by incrementing the similarity score by a predetermined value for each of the one or more anomalous machine log events that corresponds to a historical anomalous machine log event associated with the corresponding historical data elements.

3 . The service intelligence system of claim 1 , wherein the search string comprises a text search string, and the processor is configured to determine the machine identifier by applying a natural language process to the text search string.

4 . The service intelligence system of claim 1 , wherein at least one of the first set of tags or the second set of tags are generated by natural language processing of the one or more historical data elements.

5 . The service intelligence system of claim 1 , wherein at least one of the first set of historical data elements or the second set of historical data elements comprise service tickets, machine logs, part replacement logs, or a combination thereof.

6 . The service intelligence system of claim 1 , wherein the result set of corrective actions or part replacements provided in response to the search query is filtered using a domain filter, a statistical filter, or both.

7 . The service intelligence system of claim 1 , wherein the processor is configured to:
determine a maximum rank of second set of historical data elements; determine, for each historical data element of the second set of historical data elements, a weighted rank based on dividing a corresponding rank of a corresponding historical data element by the maximum rank; and select the highest ranked subset of the second set of historical data elements based on the weighted ranks.

8 . A service intelligence system, comprising:
a memory having instructions stored thereon, and a processor configured to read the instructions to:
implement a backend process, comprising:
receiving a plurality of historical data elements associated with at least one service issue and at least one corrective action or part replacement performed for at least one machine;
applying a segmentation process to each historical data element of the plurality of historical data elements to identify one or more segmented portions of each historical data element of the plurality of historical data elements;
applying a natural language process to at least one of the one or more segmented portions of each historical data element of the plurality of historical data elements to generate a plurality of tags for each historical data element of the plurality of historical data elements; and
associating, within a database, at least one historical data element of the plurality of historical data element with a first tag of the plurality of tags and one or more historical anomalous machine log events with a second tag of the plurality of tags, wherein each of the one or more historical anomalous machine log events are associated with a historical machine identifier associated with the at least one historical data element and a historical predetermined time period corresponding to a service issue represented by a corresponding historical data element of the plurality of historical data elements, and wherein the historical predetermined time period begins before the service issue occurred and ends after the service issue was resolved; and

implement a frontend process, comprising:
receiving, over a network, a search query associated with an occurrence of a device error;
search the database to identify a first set of historical data elements each associated with a service issue similar to the device error, wherein the search includes a natural language search process based on one or more terms extracted from the search query, wherein each historical data element in the first set of historical data elements is associated with the first tag;
determine a machine log identifier from the search query;
identifying one or more anomalous machine log events for a machine associated with the machine identifier for a predetermined time period;
determine a similarity score between the device error and each historical data element of the plurality of historical data elements based on a correspondence between the anomalous machine log events for the machine associated with the machine identifier and each of the one or more historical machine log events associated with each historical data element, and wherein each of the one or more historical machine log events is associated with at least one second tag;
generating a ranking of the plurality of historical data elements based on the similarity score;
generating a result set including the at least one corrective action or the at least one part replacement associated with the first tag and at least one corrective action or at least one part replacement associated with the at least one second tag associated with a highest ranked subset of the plurality of historical data elements;
removing contextual stop words from the result set by truncating a heavy tailed distribution of the result set; and
transmitting, over the network, the result set in response to the search query.

9 . The service intelligence system of claim 8 , wherein the similarity score is a value between 0 and 1, and wherein the similarity score for a corresponding historical dat

Assignees

• Siemens Medical Solutions Usa Inc

Inventors

• Daga Saurav
• Hurley Iii Robert Peter
• Morris Brian
• Saxena Priyanka

Classifications

• G06F16/243

  Natural language query formulation · CPC title

• G06Q10/063

  Operations research, analysis or management · CPC title

• G06Q10/20

  Administration of product repair or maintenance · CPC title

• G06N5/02

  Knowledge representation; Symbolic representation · CPC title

• G06F40/117

  Tagging; Marking up (details of markup languages G06F40/143); Designating a block; Setting of attributes (style sheets, e.g. eXtensible Stylesheet Language Transformation [XSLT], G06F40/154) · CPC title