Method and apparatus for situational analysis text generation

• US 8,762,134 B2
• Filed: 09/10/2013
• Issued: 06/24/2014
• Est. Priority Date: 08/30/2012
• Status: Active Grant

• Current Assignee: Arria Data2text Limited
• Original Assignee: Arria Data2text Limited
• Inventors: Reiter, Ehud B.
• Primary Examiner(s): Wozniak, James
• Assistant Examiner(s): Nguyen, Timothy
• Application Number:
  US14/023,056
• Publication Number:
  US 20140067377A1
• Time in Patent Office: 287 Days
• Field of Search: 704/9
• US Class Current: 704/9
• CPC Class Codes: G06F 40/154 Tree transformation for tree-structured or markup documents, e.g. XSLT, XSL-FO or stylesheets
• CPC Class Codes: G06F 40/169 Annotation, e.g. comment data or footnotes
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06F 40/56 Natural language generation
• CPC Class Codes: G08B 21/18 Status alarms G08B21/02 takes precedence

Abstract

Methods, apparatuses, and computer program products are described herein that are configured to generate a situational analysis text. In some example embodiments, a method is provided that comprises generating a set of messages based on one or more key events in a primary data channel and one or more significant events in one or more related data channels in response to an alert condition. The method of this embodiment may also include generating a situational analysis text based on the set of messages and the relationships between them. In some example embodiments, the situational analysis text is configured to linguistically express the one or more key events, the one or more significant events, and the relationships between the one or more key events and the one or more significant events.

Claims

1. A method for generating a situational analysis text that linguistically expresses contextual information related to an alert condition, the method comprising:

assigning at least one data channel on which the alert condition was identified as a primary data channel;

assigning one or more data channels that are identified as related to the primary data channel as one or more related data channels, wherein the one or more related data channels are a subset of one or more monitored data channels;

generating, using a processor, a set of messages based on one or more key events in the primary data channel and one or more significant events in the one or more related data channels, wherein at least one message in the set of messages corresponds to at least one of the one or more of the key events or the one or more significant events and is expressible as a simple sentence;

generating a tree-structured object that selects and orders at least a portion of the set of messages, wherein the at least the portion of the set of messages are selected based on the one or more key events and one or more significant events and are ordered based on a text schema; and

refining the tree-structured object based on domain constraints, wherein the situational analysis text is generated based on a realization of the tree-structured object; and

generating the situational analysis text based on the set of messages and a relationship between them, wherein the situational analysis text is configured to linguistically express contextual information related to the alert condition using the one or more key events, the one or more significant events, and the relationships between the one or more key events and the one or more significant events.

2. The method of claim 1 further comprising:

identifying the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

detecting one or more patterns in the primary data channel by;

identifying one or more patterns wherein a pattern is at least one of a trend, spike or step in a data channel;

assigning an importance level to the one or more patterns; and

identifying one or more key events of the one or more patterns, wherein a key event is a pattern that exceeds a predefined importance level; and

identifying one or more patterns in the one or more related data channels by;

assigning an importance level to one or more patterns; and

identifying one or more significant events of the one or more patterns, wherein a significant event is a pattern that exceeds a predefined importance level.

3. The method according to claim 2, further comprising:

determining that the one or more patterns identified in the one or more related data channels violates a predetermined constraint; and

determining that the one or more patterns are one or more unexpected patterns, wherein the one or more significant events comprise one or more patterns that violate a predetermined constraint or is determined to be an unexpected pattern.

4. The method of claim 1 further comprising:

identifying the one or more related data channels based on a domain model that provides an indication of relationships between data channels; and

detecting one or more patterns in the primary data channel by;

identifying one or more patterns wherein a pattern is at least one of a trend, spike or step in a data channel;

assigning an importance level to the one or more patterns; and

identifying one or more key events of the one or more patterns, wherein a key event is a pattern that exceeds a predefined importance level.

5. The method of claim 1 further comprising:

identifying the one or more related data channels based on a domain model that provides an indication of relationships between data channels; and

identifying one or more patterns in the one or more related data channels by;

assigning an importance level to one or more patterns; and

identifying one or more significant events of the one or more patterns, wherein a significant event is a pattern that exceeds a predefined importance level.

6. The method according to claim 1, further comprising:

identifying the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

determining that the one or more patterns identified in the one or more related data channels violates a predetermined constraint; and

determining that the one or more patterns are one or more unexpected patterns, wherein the one or more significant events comprise one or more patterns that violate a predetermined constraint or is determined to be an unexpected pattern.

7. The method according to claim 1, further comprising:

identifying the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

accessing at least one of historical data or event data for the primary data channel and the one or more related data channels; and

causing the at least one of the historical data or the event data to be expressed linguistically in the situational analysis text, wherein the historical data further comprises at least one of previous events, contextual information, background information, actions taken during a previous instance of the alert condition or results of a previous alert condition.

8. The method according to claim 1, further comprising:

identifying the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

determining one or more causes of the alert condition; and

causing one or more related data channels to be selected that correspond to the one or more causes of the alert condition, wherein the situational awareness text is further configured to linguistically express the one or more causes of the alert condition.

9. The method according to claim 1, further comprising:

identifying the one or more related data channels;

determining that a relationship exists between the primary data channel and the one or more related data channels; and

determining an expression that linguistically describes the relationship between the primary data channel and the one or more related data channels, wherein the situational awareness text is further configured to linguistically express the relationship between the primary data channel and the one or more related data channels.

10. The method according to claim 1, further comprises:

ordering the one or more messages in the tree-structured object by;

causing one or more messages relating to an alert to be included in a first portion;

causing one or more messages relating to the primary data channel in a second portion; and

causing one or more messages relating to the one or more related data channels to be included in a third portion, wherein the first portion, second portion and third portion provide a portion of the situational analysis in the situational awareness text.

11. The method according to claim 1, further comprises:

refining the tree-structured object by converting the one or more messages to one or more phrase specifications by;

determining whether to combine one or more messages in the tree structure to produce a complex sentence;

selecting one or more words to express one or more concepts in the one or more messages and the relations between them; and

determining one or more words to refer to one or more entities in at least one of the one or more messages.

12. The method according to claim 1, further comprises:

traversing the refined tree structured object to express the refined tree structured object in natural language in the form of the situational awareness text by applying a grammar to one or more phrase specifications, wherein the grammar specifies one or more valid syntactic structures in a language and provides a mapping from the one or more phrase specifications into one or more corresponding natural language sentences.

13. An apparatus configured to generate a situational analysis text that linguistically expresses contextual information related to an alert condition, the apparatus comprising:

at least one processor; and

at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least;

assign at least one data channel on which the alert condition was identified as a primary data channel;

assign one or more data channels that are identified as related to the primary data channel as one or more related data channels, wherein the one or more related data channels are a subset of one or more monitored data channels;

generate a set of messages based on one or more key events in the primary data channel and one or more significant events in one or more related data channels, wherein at least one message in the set of messages corresponds to at least one of the one or more of the key events or the one or more significant events and is expressible as a simple sentence;

generate a tree-structured object that selects and orders at least a portion of the set of messages, wherein the at least the portion of the set of messages are selected based on the one or more key events and one or more significant events and are ordered based on a text schema; and

refine the tree-structured object based on domain constraints, wherein the situational analysis text is generated based on a realization of the tree-structured object; and

generate the situational analysis text based on the set of messages and a relationship between them, wherein the situational analysis text is configured to linguistically express contextual information related to the alert condition based on the one or more key events, the one or more significant events, and the relationships between the one or more key events and the one or more significant events.

14. The apparatus of claim 13, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

identify the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

detect one or more patterns in the primary data channel by;

identifying one or more patterns wherein a pattern is at least one of a trend, spike or step in a data channel;

assigning an importance level to the one or more patterns; and

identifying one or more key events of the one or more patterns, wherein a key event is a pattern that exceeds a predefined importance level; and

identify one or more patterns in the one or more related data channels by;

assigning an importance level to one or more patterns; and

identifying one or more significant events of the one or more patterns, wherein a significant event is a pattern that exceeds a predefined importance level.

15. The apparatus according to claim 13, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

identify the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

access at least one of historical data or event data for the primary data channel and the one or more related data channels; and

cause the at least one of the historical data or the event data to be expressed linguistically in the situational analysis text, wherein the historical data further comprises at least one of previous events, contextual information, background information, actions taken during a previous instance of the alert condition or results of a previous alert condition.

16. The apparatus according to claim 13, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

identify the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

determine one or more causes of the alert condition; and

cause one or more related data channels to be selected that correspond to the one or more causes of the alert condition, wherein the situational awareness text is further configured to linguistically express the one or more causes of the alert condition.

17. The apparatus according to claim 13, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

order the one or more messages in the tree-structured object by;

causing one or more messages relating to an alert to be included in a first portion;

causing one or more messages relating to the primary data channel in a second portion; and

causing one or more messages relating to the one or more related data channels to be included in a third portion, wherein the first portion, second portion and third portion provide a portion of the situational analysis in the situational awareness text; and

refine the tree-structured object by converting the one or more messages to one or more phrase specifications by;

determining whether to combine one or more messages in the tree structure to produce a complex sentence;

selecting one or more words to express one or more concepts in the one or more messages and the relations between them; and

determining one or more words to refer to one or more entities in at least one of the one or more messages; and

traverse the refined tree structured object to express the refined tree structured object in natural language in the form of the situational awareness text by applying a grammar to one or more phrase specifications, wherein the grammar specifies one or more valid syntactic structures in a language and provides a mapping from the one or more phrase specifications into one or more corresponding natural language sentences.

18. A computer program product configured to generate a situational analysis text that linguistically expresses contextual information related to an alert condition, the computer program product comprising:

at least one computer readable non-transitory memory medium having program code instructions stored thereon, the program code instructions which when executed by an apparatus cause the apparatus at least to;

assign at least one data channel on which the alert condition was identified as a primary data channel;

assign one or more data channels that are identified as related to the primary data channel as one or more related data channels, wherein the one or more related data channels are a subset of one or more monitored data channels;

generate a set of messages based on one or more key events in the primary data channel and one or more significant events in one or more related data channels, wherein at least one message in the set of messages corresponds to at least one of the one or more of the key events or the one or more significant events and is expressible as a simple sentence;

generate a tree-structured object that selects and orders at least a portion of the set of messages, wherein the at least the portion of the set of messages are selected based on the one or more key events and one or more significant events and are ordered based on a text schema; and

refine the tree-structured object based on domain constraints, wherein the situational analysis text is generated based on a realization of the tree-structured object; and

generate the situational analysis text based on the set of messages and a relationship between them, wherein the situational analysis text is configured to linguistically express contextual information related to the alert condition based on the one or more key events, the one or more significant events, and the relationships between the one or more key events and the one or more significant events.

19. The computer program product of claim 18, further comprising program code instructions which when executed by the apparatus further cause the apparatus at least to:

identify the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

detect one or more patterns in the primary data channel by;

identifying one or more patterns wherein a pattern is at least one of a trend, spike or step in a data channel;

assigning an importance level to the one or more patterns; and

identifying one or more key events of the one or more patterns, wherein a key event is a pattern that exceeds a predefined importance level; and

identify one or more patterns in the one or more related data channels by;

assigning an importance level to one or more patterns; and

identifying one or more significant events of the one or more patterns, wherein a significant event is a pattern that exceeds a predefined importance level.

20. The computer program product according to claim 18, further comprising program code instructions which when executed by the apparatus further cause the apparatus at least to:

identify the one or more related data channels based on a domain model that provides an indication of relationships between data channels;

access at least one of historical data or event data for the primary data channel and the one or more related data channels; and

cause the at least one of the historical data or the event data to be expressed linguistically in the situational analysis text, wherein the historical data further comprises at least one of previous events, contextual information, background information, actions taken during a previous instance of the alert condition or results of a previous alert condition.