Machine-learning based transcript summarization

A system for automatically generating a summary of a multi-party interaction using natural language processing, the system comprising at least one processor and at least one memory including program code, the at least one memory and the program code configured to, with the at least one processor, cause the system to at least:

• receive a multi-party interaction transcript data object comprising a plurality of interaction utterances from at least two participants of the multi-party interaction transcript data object, wherein an interaction utterance comprises one or more sequential sentences associated with one of the at least two participants of the multi-party interaction transcript data object;
• generate, using an extractive summarization model, a key sentence of the multi-party interaction transcript data object;
• identify, from the multi-party interaction transcript data object, a particular interaction utterance of the plurality of interaction utterances that corresponds to the key sentence;
• generate a contextual summary for the multi-party interaction transcript data object based at least in part on the particular interaction utterance;
• and
• generate, using a machine-learning based speech converter model, a reported contextual summary based at least in part on the contextual summary, wherein the reported contextual summary comprises the contextual summary from a perspective of a particular participant.

A computer program product for generating a summary of a multi-party interaction using natural language processing, the computer program product comprising at least one non-transitory computer-readable storage medium having computer-readable program code portions stored therein that, when executed by one or more processors, cause the one or more processors to:

• receive a multi-party interaction transcript data object comprising a plurality of interaction utterances from at least two participants of the multi-party interaction transcript data object, wherein an interaction utterance comprises one or more sequential sentences associated with one of the at least two participants of the multi-party interaction transcript data object;
• generate, using an extractive summarization model, a key sentence of the multi-party interaction transcript data object;
• identify, from the multi-party interaction transcript data object, a particular interaction utterance of the plurality of interaction utterances that corresponds to the key sentence;
• generate a contextual summary for the multi-party interaction transcript data object based at least in part on the particular interaction utterance;
• and
• generate, using a machine-learning based speech converter model, a reported contextual summary based at least in part on the contextual summary, wherein the reported contextual summary comprises the contextual summary from a perspective of a particular participant.