Query modification based on non-textual resource context
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, modifying queries based on non-textual content. In one aspect, a method includes receiving, from a user device, a query including a plurality of terms; determining active non-textual data displayed in an application environment on the user device; determining, from the non-textual textual data, modification data for the query; generating a set of modified queries based on the query and the modification parameters; scoring the modified queries according to one or more scoring criteria; selecting one of the modified queries based on the scoring; and providing, to the user device, search results responsive to the selected modified query.
1 . A method implemented by one or more processors, the method comprising:
receiving a plurality of terms spoken by a user of a user device, the user device displaying a plurality of non-textual data elements in an application environment when the user spoke the plurality of terms;
determining that the plurality of terms is an ambiguous instruction, and in response:
identifying, as an active non-textual data element, one of the non-textual data elements that has been selected by the user;
automatically determining, by the data processing apparatus, for the active non-textual data element, modification data from the active non-textual data element for the ambiguous instruction;
automatically generating, by the data processing apparatus, a modified instruction based on the ambiguous instruction and the modification data; and
providing, to the user device, the modified instruction, wherein the modified instruction causes the user device to perform an action defined by the modified instruction.
2 . The method of claim 1 , wherein the active non-textual data element comprises an image that is selected by the user.
3 . The method of claim 1 , wherein the active non-textual data element comprises a video that is selected by the user.
4 . The method of claim 1 , wherein the active non-textual data element comprises an image that is captured by a camera of the user device.
5 . The method of claim 1 , wherein determining modification data for the ambiguous instruction comprises:
determining, from the active non-textual data element, one or more labels that describe subject matter of the active non-textual data element; and
for each label, generating a modified instruction based on the ambiguous instruction and the label.
6 . The method of claim 5 , wherein generating a modified instruction based on the ambiguous instruction and the label comprises concatenating the ambiguous instruction with the label.
7 . The method of claim 1 , wherein determining modification data for the ambiguous instruction comprises:
determining, from the active non-textual data element, entity text describing one or more entities associated with the active non-textual data element; and
for each entity, generating a modified instruction based on the ambiguous instruction and the entity.
8 . The method of claim 7 , wherein generating a modified instruction based on the ambiguous instruction and the entity text describing the entity comprises revising one or more terms of the ambiguous instruction based on the entity text describing the entity.
9 . The method of claim 1 , wherein receiving a plurality of terms spoken by a user of a user device comprises receiving text generated by a speech analyzer.
10 . A system comprising one or more processors and memory storing instructions that, in response to execution by the one or more processors, cause the one or more processors to:
receive a plurality of terms spoken by a user of a user device, the user device displaying a plurality of non-textual data elements in an application environment when the user spoke the plurality of terms;
determine that the plurality of terms is an ambiguous instruction, and in response:
identify, as an active non-textual data element, one of the non-textual data elements that has been selected by the user;
automatically determine, by the data processing apparatus, for the active non-textual data element, modification data from the active non-textual data element for the ambiguous instruction;
automatically generate, by the data processing apparatus, a modified instruction based on the ambiguous instruction and the modification data; and
provide, to the user device, the modified instruction, wherein the modified instruction causes the user device to perform an action defined by the modified instruction.
11 . The system of claim 10 , wherein the active non-textual data element comprises an image that is selected by the user.
12 . The system of claim 10 , wherein the active non-textual data element comprises a video that is selected by the user.
13 . The system of claim 10 , wherein the active non-textual data element comprises an image that is captured by a camera of the user device.
14 . The system of claim 10 , wherein the instructions to determine modification data for the ambiguous instruction comprise instructions to:
determine, from the active non-textual data element, one or more labels that describe subject matter of the active non-textual data element; and
for each label, generate a modified instruction based on the ambiguous instruction and the label.
15 . The system of claim 14 , wherein the instructions to generate a modified instruction based on the ambiguous instruction and the label comprise instructions to concatenate the ambiguous instruction with the label.
16 . The system of claim 10 , wherein the instructions to determine modification data for the ambiguous instruction comprise instructions to:
determine, from the active non-textual data element, entity text describing one or more entities associated with the active non-textual data element; and
for each entity, generate a modified instruction based on the ambiguous instruction and the entity.
17 . The system of claim 16 , wherein the instructions to generate a modified instruction based on the ambiguous instruction and the entity text describing the entity comprise instructions to revise one or more terms of the ambiguous instruction based on the entity text describing the entity.
18 . The system of claim 10 , wherein the instructions to receive a plurality of terms spoken by a user of a user device comprise instructions to receive text generated by a speech analyzer.
19 . At least one non-transitory computer-readable medium comprising instructions that, in response to execution by one or more processors, cause the one or more processors to:
receive a plurality of terms spoken by a user of a user device, the user device displaying a plurality of non-textual data elements in an application environment when the user spoke the plurality of terms;
determine that the plurality of terms is an ambiguous instruction, and in response:
identify, as an active non-textual data element, one of the non-textual data elements that has been selected by the user;
automatically determine, by the data processing apparatus, for the active non-textual data element, modification data from the active non-textual data element for the ambiguous instruction;
automatically generate, by the data processing apparatus, a modified instruction based on the ambiguous instruction and the modification data; and
provide, to the user device, the modified instruction, wherein the modified instruction causes the user device to perform an action defined by the modified instruction.
20 . The at least one non-transitory computer-readable medium of claim 19 , wherein the active non-textual data element comprises an image that is selected by the user.