Assessment of inconsistent statements to a recipient group

A computer program product comprising a computer readable storage medium having program code embodied therewith, the program code executable by a processor to:

• receive a statement targeting a first recipient group;
• access a corpus of historical statements, the corpus of historical statements comprising:
• at least one first historical statement addressed to the first recipient group;
• at least one second historical statement addressed to a second recipient group not targeted by the received statement;
• and
• optionally at least one third historical statement addressed to a third recipient group, at least one recipient of which is not targeted by the received statement;
• determine a match between the first recipient group targeted by the received statement and the first recipient group to which the at least one first historical statement is addressed and select the at least one first historical statement based on the match;
• determine an absence of a match between the first recipient group targeted by the received statement and the second recipient group to which the at least one second historical statement is addressed and disregard the at least one second historical statement based on the absence of the match;
• optionally determine that the third recipient group satisfies a commonality threshold with the first recipient group and optionally select the third historical statement based on satisfaction of the commonality threshold;
• optionally identify at least one historical cross-statement between the first recipient group and the third recipient group satisfying the commonality threshold and optionally select the at least one historical cross-statement;
• apply natural language processing (NLP) to the received statement, the selected at least one first historical statement, optionally the at least one third historical statement, and optionally the at least one historical cross-statement to generate corresponding content data;
• assess the content data corresponding to the received statement for an inconsistency with the content data corresponding to the selected at least one first historical statement, the optionally selected at least one third historical statement, the optionally selected at least one historical cross-statement, or any combination thereof;
• and
• provide a notification of the inconsistency.

A machine-implemented method comprising:

• receiving a statement of an entity targeting a first recipient group;
• accessing, by a processor of a computer device, a corpus of historical statements, the corpus of historical statements comprising:
• at least one first historical statement addressed to the first recipient group;
• at least one second historical statement addressed to a second recipient group not targeted by the received statement;
• and
• optionally at least one third historical statement addressed to a third recipient group, at least one recipient of which is not targeted by the received statement;
• determining, by the processor, a match between the first recipient group targeted by the received statement and the first recipient group to which the at least one first historical statement is addressed and selecting the at least one first historical statement based on the match;
• determining, by the processor, an absence of a match between the first recipient group targeted by the received statement and the second recipient group to which the at least one second historical statement is addressed and disregarding the at least one second historical statement based on the absence of the match;
• optionally determining, by the processor, that the third recipient group satisfies a commonality threshold with the first recipient group and optionally selecting the third historical statement based on satisfaction of the commonality threshold;
• optionally identifying, by the processor, at least one historical cross-statement between the first recipient group and the third recipient group satisfying the commonality threshold and optionally selecting the at least one historical cross-statement;
• applying natural language processing (NLP) to the received statement, the selected at least one first historical statement, optionally the at least one third historical statement, and optionally the at least one historical cross-statement to generate corresponding content data;
• applying, by the processor, the content data to an artificial intelligence (AI) platform, the applying comprising assessing the content data corresponding to the received statement for an inconsistency with the content data corresponding to the selected at least one first historical statement, the optionally selected at least one third historical statement, the optionally selected at least one historical cross-statement, or any combination thereof;
• and
• providing, by the processor, a notification of the inconsistency.