Back to results
US 11,689,432 B1Granted
Feedback prioritization system
AssigneeAmazon Technologies, Inc.
InventorsKasturi Bhattacharjee, Rashmi Gangadharaiah, Sharon Shapira, Ankit Kapoor, Tony Chun Tung Ng, Senthil Chock Chidambaram, Deepak Seetharam Nadig
Life of this patent
Estimating the term…
Abstract
The present disclosure generally relates to a feedback processing service that can receive customer input, as customer feedback corresponds to a service context. The feedback processing service aggregates semantically similar feedback as a cluster. Then, the feedback processing service can prioritize each of the clusters by ranking each of the clusters.
Claims33
A system for prioritizing customers' feedback in network-based services, the system comprising
- one or more computing devices associated with a processor and a memory for executing computer-executable instructions to implement a feedback management service, wherein the feedback management service is configured to:
- obtain a set of customer inputs as customer feedback related to one or more network services hosted on behalf of network service providers;
- vectorize the set of customer inputs to form a set of vectorized customer inputs, wherein vectorizing the set of customer inputs includes generating an individual numerical vector for individual customer inputs from the set of customer inputs;
- in the control policy input includes a specification of a network service type associated with the plurality of network services;
- cluster the set of vectorized customer inputs, wherein clustering the set of vectorized inputs includes:
- aggregate semantically similar vectorized customer inputs based on comparison of the numerical vectors;
- and
- for individual clusters, identify a representative vectorized customer input based on a centroid of the individual cluster;
- for the individual cluster, filter any cluster not exceeding a minimal vectorized customer input threshold to form a set of filtered clusters;
- for the filtered clusters, prioritize the individual cluster, wherein the prioritization comprising;
- measuring a ranking criteria;
- and
- ranking the cluster based on the measured criteria;
- and
- generate a processing result based on the prioritized, filtered clusters.
12 more elements
A method for managing network-based feedback processing service comprising:
- vectorizing a set of customer input associated with a network service to identify customer feedbacks for a plurality of network services hosted on behalf of network service providers;
- clustering the set of vectorized customer inputs, wherein clustering the set of vectorized customer inputs comprise:
- aggregating semantically similar vectorized customer inputs based on comparison of the numerical vectors;
- and
- for individual clusters, identifying a representative vectorized customer input based on a centroid of the individual cluster;
- prioritizing the individual clusters, wherein the prioritization comprises:
- measuring a ranking criteria;
- and
- ranking the cluster based on the measured criteria;
- and
- generating a processing result based on the prioritized clusters.
8 more elements
A method for managing network-based feedback processing service comprising:
- obtaining a set of customer inputs as customer feedback related to one or more network services hosted on behalf of network service providers, wherein the set of customer inputs includes textual inputs characterized as customer feedback;
- vectorizing the set of customer inputs to form a set of vectorized customer inputs, wherein vectorizing the set of customer inputs includes generating an individual numerical vector for individual customer inputs from the set of customer inputs;
- in the control policy input includes a specification of a network service type associated with the plurality of network services;
- clustering the set of vectorized customer inputs, wherein clustering the set of vectorized customer inputs comprise:
- aggregating semantically similar vectorized customer inputs based on comparison of the numerical vectors, wherein the aggregation uses k-means clustering, wherein the k-means clustering can use a sentence-bert (“SBERT”) technique;
- and
- for individual clusters, identifying a representative vectorized customer input based on a centroid of the individual cluster;
- for the individual cluster, filter any cluster not exceeding a minimal vectorized customer input threshold to form a set of filtered clusters;
- for the filtered clusters, prioritize the individual cluster, wherein the prioritization comprises:
- measuring an inertia score of each of the clusters, wherein the inertia score represents a coherence of each of the clusters.;
- and
- ranking the cluster in ascending order of the inertia score;
- and
- generating processing results based on the prioritized clusters.
11 more elements
Dependencies are inferred from the claim text.