ML Issue Mapping From Customer Dialogues for Streaming QoE Gaps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional quality of experience (QoE) metrics in video streaming services fail to identify user-centric technical issues on client devices, such as non-functional pause buttons and audio-video synchronization problems, which remain unresolved due to their system-centric focus.
Innovation Solution
Training machine learning models to analyze user-derived natural language inputs, particularly from customer service interactions, to identify and categorize technical issues, providing actionable insights and potential resolutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If system-centric QoE metrics are used to monitor video streaming services, then data transmission aspects such as rebuffer rate and error rates can be measured, but user-centric technical issues on client devices remain unobservable and unresolved
Solution Approach 1:
Instead of relying on server-side system metrics to infer user experience, the patent inverts the approach by directly capturing user-derived natural language inputs from customer service interactions. This inversion allows the system to observe client-side issues directly from user reports rather than inferring them from transmission data, thereby expanding the scope of detectable technical problems while maintaining measurement precision.
Solution Approach 2:
The patent introduces machine learning models as an intermediary between raw customer service transcripts and technical issue identification. This intermediary processes unstructured natural language dialogue, extracts relevant technical information, and structures it for analysis, enabling the system to bridge the gap between user complaints and actionable technical insights without requiring direct access to client device metrics.
2Loss of information
If machine learning models are trained to analyze natural language dialogue from customer interactions, then user-centric technical issues can be identified, but additional data processing complexity is introduced
Solution Approach 1:
The patent segments the data processing pipeline into distinct machine learning model components: one model for generating structured data from natural language transcripts, another for aggregating structured data into clusters, and a third for mapping clusters to technical issues. This segmentation allows each component to specialize in a specific task, improving information extraction effectiveness while making the overall complex system more manageable and maintainable.
Solution Approach 2:
The machine learning models are trained to autonomously process and interpret customer service transcripts without requiring manual annotation or intervention. The models self-learn from historical data, automatically identifying patterns and extracting technical issue information, thereby reducing the need for human labor in data processing while maintaining high visibility of user-centric issues.
3Reliability
If natural language dialogue data is processed to identify technical issues, then actionable insights can be obtained, but time and computational resources are consumed
Solution Approach 1:
The patent implements preliminary action by pre-training machine learning models on historical customer service transcripts and technical issue data before deployment. This offline training phase allows the models to learn patterns and relationships in advance, so that during actual operation, they can quickly process new transcripts and identify technical issues with high accuracy without consuming excessive processing time during critical service periods.
Solution Approach 2:
The patent replaces manual analysis of customer service transcripts with automated machine learning models. This substitution eliminates the need for human analysts to manually review and categorize each transcript, dramatically reducing processing time while maintaining or improving identification accuracy through consistent application of learned patterns across all data.
Data Source
AI summary
A computer-implemented method includes accessing natural language dialogue data gathered from interactions between an entity and a user. The method further includes accessing contextual information related to the user, which provides additional indications of the user's consumption of the service. The method further includes training a machine learning model to generate structured data that is assembled according to a schema that prepares the structured data for aggregation into clusters, training the machine learning model to aggregate the data into clusters, where each cluster includes an identified standard that defines inclusion in the cluster, training the machine learning model to map the aggregated, structured data to a specified technical issue related to the service that is to be resolved and, based on the mapping, providing an indication of the specified technical issue that is to be resolved to a computing device. Various other methods, systems, and computer-readable media are also disclosed.