MIXED-GRAINED DETECTION AND ANALYSIS OF USER LIFE EVENTS FOR CONTEXT UNDERSTANDING
Techniques for resolving multiple user requests from multiple user accounts by an interactive interface are described. An interactive interface can obtain a first multi-dimensional context graph for a first user account and a second context graph for a second user account. Each graph comprises correlated contexts related to the user account. The interface can also receive a first user request associated with the first user account and a second user request associated with the second user account; determine, based on the first graph, a first current context and one or more first previous contexts for the first user request; determine, based on the second graph, a second current context and one or more second previous contexts for the second user request; determine one or more interrelationships between the first and the second graphs; and resolve the user requests based on the contexts and the interrelationships.
1 . A computer-implemented method for resolving multiple user requests from a user account by an interactive interface, comprising:
obtaining, by a processor, a multi-dimensional context graph associated with a user account, wherein the context graph comprises correlated contexts related to the user account, wherein each context is represented as node in the context graph, wherein each correlation between two contexts is represented as an edge in the context graph, and wherein the context graph includes one or more context clusters, and wherein each correlation between two context clusters is represented as an interrelationship in the context graph;
receiving a first user request and a second user request, each of the requests being associated with the user account;
for each of the first and second user requests:
determining, based on the context graph, a current context for the user request and one or more first previous contexts that are related to the current context, wherein the current context and the first previous contexts are part of a same context cluster;
determining, based on the context graph, one or more second previous contexts that are related to the current context, wherein the current context and the second previous contexts are not part of the same context cluster; and
resolving the user request based on the current context, the one or more first previous contexts, and the one or more second previous contexts.
2 . The method of claim 1 , further comprising updating the one or more interrelationships between context clusters as data associated with the user account evolves.
3 . The method of claim 2 , wherein determining the one or more first previous contexts includes determining a context cluster to which the current context belongs.
4 . The method of claim 3 , further comprising determining, for each current cluster, that the current context's cluster is related to another context cluster based on one or more interrelationships between the current context's cluster and the other cluster.
5 . The method of claim 3 , further comprising determining, for each current cluster, that the current context's cluster is related to another context cluster based on one or more edges within the current context's cluster.
6 . The method of claim 1 , wherein none of the user requests is resolved prior to receipt of the other user request.
7 . The method of claim 1 , wherein determining each of the current contexts is based, at least in part, on determining one or more key identifiers within the user requests.
8 . A non-transitory computer readable medium comprising instructions for resolving multiple user requests from a user account by an interactive interface, which when executed by one or more processors, cause the one or more processors to:
obtain a context graph associated with a user account, wherein the context graph comprises correlated contexts related to the user account, wherein each context is represented as node in the context graph, wherein each correlation between two contexts is represented as an edge in the context graph, and wherein the context graph includes one or more context clusters, and wherein each correlation between two context clusters is represented as an interrelationship in the context graph;
receive a first user request and a second user request, each of the requests being associated with the user account;
for each of the first and second user requests:
determine, based on the context graph, a current context for the user request and one or more first previous contexts that are related to the current context, wherein the current context and the first previous contexts are part of a same context cluster;
determine, based on the context graph, one or more second previous contexts that are related to the current context, wherein the current context and the second previous contexts are not part of the same context cluster; and
resolve the user request based on the current context, the one or more first previous contexts, and the one or more second previous contexts.
9 . The non-transitory computer readable medium of claim 8 , further comprising instructions for causing the one or more processors to update the one or more interrelationships between context clusters as data associated with the user account evolves.
10 . The non-transitory computer readable medium of claim 9 , wherein the instructions for causing the one or more processors to determine the one or more first previous contexts includes instructions for causing the one or more processors to determine a context cluster to which the current context belongs.
11 . The non-transitory computer readable medium of claim 10 , further comprising instructions for causing the one or more processors to determine, for each current cluster, that the current context's cluster is related to another context cluster based on one or more interrelationships between the current context's cluster and the other cluster.
12 . The non-transitory computer readable medium of claim 10 , further comprising instructions for causing the one or more processors to determine, for each current cluster, that the current context's cluster is related to another context cluster based on one or more edges within the current context's cluster.
13 . The non-transitory computer readable medium of claim 8 , wherein none of the user requests is resolved prior to receipt of the other user request.
14 . The non-transitory computer readable medium of claim 8 , wherein the instructions for causing the one or more processors to determine each of the current contexts is based, at least in part, on instructions that cause the one or more processors to determine one or more key identifiers within the user requests.
15 . A processing system configured to resolve multiple user requests from multiple user accounts, comprising:
logic configured to:
obtain a context graph associated with a user account, wherein the context graph comprises correlated contexts related to the user account, wherein each context is represented as node in the context graph, wherein each correlation between two contexts is represented as an edge in the context graph, and wherein the context graph includes one or more context clusters, and wherein each correlation between two context clusters is represented as an interrelationship in the context graph;
receive a first user request and a second user request, each of the requests being associated with the user account;
for each of the first and second user requests:
determine, based on the context graph, a current context for the user request and one or more first previous contexts that are related to the current context, wherein the current context and the first previous contexts are part of a same context cluster;
determine, based on the context graph, one or more second previous contexts that are related to the current context, wherein the current context and the second previous contexts are not part of the same context cluster; and
resolve the user request based on the current context, the one or more first previous contexts, and the one or more second previous contexts.
16 . The system of claim 15 , wherein the logic is further configured to update the one or more interrelationships between context clusters as data associated with the user account evolves.
17 . The system of claim 16 , wherein the logic being configured to determine the one or more first previous contexts includes the logic being configured to determine a context cluster to which the current context belongs.
18 . The system of claim 17 , wherein the logic is further configured to determine, for each current cluster, that the current context's cluster is related to another context cluster based on one or more interrelationships between the current context's cluster and the other cluster.
19 . The system of claim 17 , wherein the logic is further configured to determine, for each current cluster, that the current context's cluster is related to another context cluster based on one or more edges within the current context's cluster.
20 . The system of claim 15 , wherein the logic being configured to determine each of the current contexts is based, at least in part, on the logic being configured to determine one or more key identifiers within the user requests.