Multi-attribute evaluation of narratives using event datasets

A method comprising:

• in a system comprising one or more processor devices operatively coupled to one or more memory devices, the one or more processor devices executing steps comprising:
• obtaining input representing at least one historical dataset of events;
• obtaining input representing a set of candidate narratives, wherein each candidate narrative is a potential future event sequence;
• obtaining input representing domain knowledge, wherein the domain knowledge is associated with at least a portion of the at least one historical dataset of events;
• obtaining input representing a query, wherein the query comprises one or more events of interest to a user;
• computing attribute scores for one or more first candidate narratives based on at least a portion of the obtained inputs;
• and
• creating one or more second candidate narratives, the one or more second candidate narratives being different from the one or more first candidate narratives;
• and
• wherein one of the attribute scores comprises a plausibility attribute score representing a measure estimating a likelihood that a given candidate narrative will occur in the future;
• wherein a second one of the attribute scores comprises a surprise attribute score representing a measure estimating how surprising a given candidate narrative will be to the user;
• wherein a third one of the attribute scores comprises an impact attribute score representing a measure estimating how impactful a given candidate narrative will be based on at least a portion of the domain knowledge;
• wherein computing attribute scores includes computing attribute scores for the one or more second candidate narratives;
• wherein the attribute scores for the one or more first candidate narratives represent a first likelihood of an occurrence of one or more first potential future event sequences;
• wherein the attribute scores for the one or more second candidate narratives represent a second likelihood of an occurrence of one or more second potential future event sequences;
• and
• wherein, based at least in part on the respective attribute scores, the second likelihood of an occurrence of the one or more second potential future events sequences is less than the first likelihood of an occurrence of the one or more first potential future events sequences.

An article of manufacture comprising a processor-readable storage medium having encoded therein executable code of one or more software programs, wherein the one or more software programs when executed cause one or more processor devices operatively coupled to one or more memory devices to perform steps of:

• obtaining input representing at least one historical dataset of events;
• obtaining input representing a set of candidate narratives, wherein each candidate narrative is a potential future event sequence;
• obtaining input representing domain knowledge, wherein the domain knowledge is associated with at least a portion of the at least one historical dataset of events;
• obtaining input representing a query, wherein the query comprises one or more events of interest to a user;
• computing attribute scores for one or more first candidate narratives based on at least a portion of the obtained inputs;
• and
• creating one or more second candidate narratives, the one or more second candidate narratives being different from the one or more first candidate narratives;
• and
• wherein one of the attribute scores comprises a plausibility attribute score representing a measure estimating a likelihood that a given candidate narrative will occur in the future;
• wherein a second one of the attribute scores comprises a surprise attribute score representing a measure estimating how surprising a given candidate narrative will be to the user;
• wherein a third one of the attribute scores comprises an impact attribute score representing a measure estimating how impactful a given candidate narrative will be based on at least a portion of the domain knowledge;
• wherein computing attribute scores includes computing attribute scores for the one or more second candidate narratives;
• wherein the attribute scores for the one or more first candidate narratives represent a first likelihood of an occurrence of one or more first potential future event sequences;
• wherein the attribute scores for the one or more second candidate narratives represent a second likelihood of an occurrence of one or more second potential future event sequences;
• and
• wherein, based at least in part on the respective attribute scores, the second likelihood of an occurrence of the one or more second potential future events sequences is less than the first likelihood of an occurrence of the one or more first potential future events sequences.