Disambiguating user input with memorization for improved user assistance

One or more computer-readable non-transitory storage media embodying software that is operable when executed to:

• receive, from a client system associated with a first user during a first dialog session, a user input by the first user, wherein the user input comprises an ambiguous mention;
• analyze, based on a natural-language understanding module, the user input to identify one or more intents and one or more slots, wherein at least one of the one or more slots is an ambiguous slot corresponding to the ambiguous mention;
• identify one or more candidate entities corresponding to the ambiguous slot;
• determine one or more tasks based on the one or more intents associated with the user input;
• rank the one or more candidate entities based on user profile data associated with the first user;
• send, to the client system associated with the first user during the first dialog session, instructions for prompting the first user to select a candidate entity from a set of candidate entities having a rank greater than a threshold rank, wherein the selected candidate entity is resolved to the ambiguous slot;
• receive, from the client system during the first dialog session, a user selected candidate entity from the first user;
• determine, during the first dialog session, whether to store the user selected candidate entity together with the ambiguous mention as a mention-entity pair in a dialog state associated with the first dialog session or as part of the user profile data based on whether the stored mention-entity pair is to be used for resolving the ambiguous mention during the first dialog session or across a plurality of dialog sessions;
• and
• store, the mention-entity pair during the first dialog session, wherein the mention-entity pair is associated with the one or more tasks and a decay factor indicating a confidence for resolving the ambiguous mention based on the mention-entity pair when the mention-entity pair is stored as part of the user profile data, wherein the decay factor is determined based on a slot-type of the ambiguous slot corresponding to the ambiguous mention and the one or more intents, and wherein the mention-entity pair is discarded after the first dialog session when the mention-entity pair is stored in the dialog state associated with the first dialog session.

A system comprising:

• one or more processors;
• and a non-transitory memory coupled to the processors comprising instructions executable by the processors, the processors operable when executing the instructions to:
• receive, from a client system associated with a first user during a first dialog session, a user input by the first user, wherein the user input comprises an ambiguous mention;
• analyze, based on a natural-language understanding module, the user input to identify one or more intents and one or more slots, wherein at least one of the one or more slots is an ambiguous slot corresponding to the ambiguous mention;
• identify one or more candidate entities corresponding to the ambiguous slot;
• determine one or more tasks based on the one or more intents associated with the user input;
• rank the one or more candidate entities based on user profile data associated with the first user;
• send, to the client system associated with the first user during the first dialog session, instructions for prompting the first user to select a candidate entity from a set of candidate entities having a rank greater than a threshold rank, wherein the selected candidate entity is resolved to the ambiguous slot;
• receive, from the client system during the first dialog session, a user selected candidate entity from the first user;
• determine, during the first dialog session, whether to store the user selected candidate entity together with the ambiguous mention as a mention-entity pair in a dialog state associated with the first dialog session or as part of the user profile data based on whether the stored mention-entity pair is to be used for resolving the ambiguous mention during the first dialog session or across a plurality of dialog sessions;
• and
• store, the mention-entity pair during the first dialog session, wherein the mention-entity pair is associated with the one or more tasks and a decay factor indicating a confidence for resolving the ambiguous mention based on the mention-entity pair when the mention-entity pair is stored as part of the user profile data, wherein the decay factor is determined based on a slot-type of the ambiguous slot corresponding to the ambiguous mention and the one or more intents, and wherein the mention-entity pair is discarded after the first dialog session when the mention-entity pair is stored in the dialog state associated with the first dialog session.