Disambiguating user intent in conversational interaction system for large corpus information retrieval
View Patent ↗A method of disambiguating user intent in conversational interactions for information retrieval is disclosed. The method includes providing access to a set of content items with metadata describing the content items and providing access to structural knowledge showing semantic relationships and links among the content items. The method further includes providing a user preference signature, receiving a first input from the user that is intended by the user to identify at least one desired content item, and determining an ambiguity index of the first input. If the ambiguity index is high, the method determines a query input based on the first input and at least one of the structural knowledge, the user preference signature, a location of the user, and the time of the first input and selects a content item based on comparing the query input and the metadata associated with the content item.
1. A method of disambiguating user intent in conversational interactions for information retrieval, the method comprising:
receiving information for a set of content items, each of the content items being associated with metadata that describes the corresponding content items;
receiving information for structural knowledge, the structural knowledge showing relationships of multiple types among the content items;
receiving information for a user preference signature, the user preference signature describing preferences of a user for at least one of (i) particular content items and (ii) metadata associated with the content items;
receiving a first input from the user, the first input intended by the user to identify at least one desired content item;
determining an ambiguity index of the first input;
upon a condition in which the ambiguity index exceeds a first threshold value:
upon a condition in which at least one of the structural knowledge, the user preference signature, a location of the user, and a time of the first input is available:
determining a disambiguated first input based on the first input and the at least one of the structural knowledge, the user preference signature, the location of the user, and the time of the first input;
upon a condition in which the at least one of the structural knowledge, the user preference signature, the location of the user, and the time of the first input is not available:
presenting a question to the user;
receiving a second input from the user in response to the question; and
determining a disambiguated first input based on the first input and the second input; and
selecting a subset of content items from the set of content items based on comparing the disambiguated first input and the metadata associated with the subset of content items; and
upon a condition in which the ambiguity index does not exceed the first threshold value, selecting a subset of content items from the set of content items based on comparing the first input and the metadata associated with the subset of content items.
2. The method of claim 1 , further comprising presenting the subset of content items to the user.
3. The method of claim 1 , wherein the ambiguity index is determined based on a number of possible interpretations of the first input.
4. The method of claim 1 , wherein determining the disambiguated first input further comprises determining which portion of the first input is ambiguous, wherein the determination of the disambiguated first input is further based on the ambiguous portion of the first input.
5. The method of claim 1 , wherein determining the disambiguated first input further comprises determining intent, entity, and filter of the first input, the intent being what is sought by the user, the entity being a noun or pronoun describing the intent, and the filter being a qualifier of the entity.
6. A system for disambiguating user intent in conversational interactions for information retrieval, the system comprising:
a processor configured to:
receive information for a set of content items, each of the content items being associated with metadata that describes the corresponding content items;
receive information for structural knowledge, the structural knowledge showing relationships of multiple types among the content items;
receive information for a user preference signature, the user preference signature describing preferences of a user for at least one of (i) particular content items and (ii) metadata associated with the content items;
receive a first input from the user, the first input intended by the user to identify at least one desired content item;
determine an ambiguity index of the first input;
upon a condition in which the ambiguity index exceeds a first threshold value:
upon a condition in which at least one of the structural knowledge, the user preference signature, a location of the user, and a time of the first input is available:
determine a disambiguated first input based on the first input and the at least one of the structural knowledge, the user preference signature, the location of the user, and the time of the first input;
upon a condition in which the at least one of the structural knowledge, the user preference signature, the location of the user, and the time of the first input is not available:
present a question to the user;
receive a second input from the user in response to the question; and
determine a disambiguated first input based on the first input and the second input; and
select a subset of content items from the set of content items based on comparing the disambiguated first input and the metadata associated with the subset of content items; and
upon a condition in which the ambiguity index does not exceed the first threshold value, select a subset of content items from the set of content items based on comparing the first input and the metadata associated with the subset of content items.
7. The system of claim 6 , wherein the processor is further configured to present the subset of content items to the user.
8. The system of claim 6 , wherein the ambiguity index is determined based on a number of possible interpretations of the first input.
9. The system of claim 6 , wherein determining the disambiguated first input further comprises determining which portion of the first input is ambiguous, wherein the determination of the disambiguated first input is further based on the ambiguous portion of the first input.
10. The system of claim 6 , wherein determining the disambiguated first input further comprises determining intent, entity, and filter of the first input, the intent being what is sought by the user, the entity being a noun or pronoun describing the intent, and the filter being a qualifier of the entity.
11. A non-transitory computer-readable medium storing a plurality of instructions for controlling a data processor for disambiguating user intent in conversational interactions for information retrieval, the plurality of instructions including:
an instruction that causes the data processor to receive information for a set of content items, each of the content items being associated with metadata that describes the corresponding content items;
an instruction that causes the data processor to receive information for structural knowledge, the structural knowledge showing relationships of multiple types among the content items;
an instruction that causes the data processor to receive information for a user preference signature, the user preference signature describing preferences of a user for at least one of (i) particular content items and (ii) metadata associated with the content items;
an instruction that causes the data processor to receive a first input from the user, the first input intended by the user to identify at least one desired content item;
an instruction that causes the data processor to determine an ambiguity index of the first input;
upon a condition in which the ambiguity index exceeds a first threshold value:
upon a condition in which at least one of the structural knowledge, the user preference signature, a location of the user, and a time of the first input is available:
an instruction that causes the data processor to determine a disambiguated first input based on the first input and the at least one of the structural knowledge, the user preference signature, the location of the user, and the time of the first input;
upon a condition in which the at least one of the structural knowledge, the user preference signature, the location of the user, and the time of the first input is not available:
an instruction that causes the data processor to present a question to the user;
an instruction that causes the data processor to receive a second input from the user in response to the question; and
an instruction that causes the data processor to determine a disambiguated first input based on the first input and the second input; and
an instruction that causes the data processor to select a subset of content items from the set of content items based on comparing the disambiguated first input and the metadata associated with the subset of content items; and
upon a condition in which the ambiguity index does not exceed the first threshold value, an instruction that causes the data processor to select a subset of content items from the set of content items based on comparing the first input and the metadata associated with the subset of content items.