Feedback Report Clustering via Term Relevance Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems lack an efficient method to cluster and analyze feedback reports from computing devices, particularly in identifying relevant terms and assigning reports to appropriate clusters based on relevance, which hinders effective error reporting and analysis.
Innovation Solution
The method involves receiving feedback reports, identifying relevant terms, calculating term and resource relevance values, and implementing a clustering process to assign reports to clusters based on these values, using techniques like TF-IDF and sharding to increase efficiency and scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If feedback reports are collected and analyzed manually or using basic methods, then analysis accuracy may be maintained, but productivity and efficiency of error identification deteriorate due to the large volume of reports
Solution Approach 1:
The patent segments the large volume of feedback reports into meaningful clusters based on term relevance and similarity. By dividing the monolithic set of reports into smaller, organized clusters, the system enables efficient processing and analysis of individual clusters rather than dealing with the entire dataset at once, thereby improving productivity while managing the quantity of reports.
Solution Approach 2:
The system changes parameters by calculating term relevance values and using these to organize reports. By transforming the raw report data into a structured format with computed relevance metrics, the system enables automated clustering and efficient retrieval, significantly improving analysis efficiency without being constrained by the volume of reports.
2Loss of information
If all feedback reports are processed and stored in detail, then completeness of information is improved, but device complexity and computational resources required deteriorate
Solution Approach 1:
The patent extracts key terms and essential information from feedback reports to create term relevance values and cluster identifiers. By taking out only the most relevant features and organizing reports based on these extracted elements, the system maintains the essential information needed for analysis while reducing the overall complexity of processing and storing complete report details.
Solution Approach 2:
The system performs preliminary actions by pre-calculating term relevance values and organizing reports into clusters before full analysis is needed. This advance organization reduces the complexity of subsequent analysis operations, as the heavy lifting of structuring the data is completed beforehand, allowing faster and simpler retrieval and analysis when needed.
3Measurement precision
If term relevance threshold is set low to include more reports in clustering, then clustering completeness is improved, but processing time and computational resources worsen
Solution Approach 1:
The patent applies partial action by using a term relevance threshold that filters reports to include only those meeting the relevance criteria. Rather than processing all reports regardless of relevance, the system selectively processes only the necessary subset, thereby maintaining measurement precision while reducing processing time and computational resources required for the clustering operation.
4Device complexity
If feedback reports are clustered without automated relevance calculation, then system simplicity is maintained, but analysis accuracy and error detection capability deteriorate
Solution Approach 1:
The system implements self-service by automatically calculating term relevance values and performing clustering without requiring manual intervention or complex external analysis tools. This automated approach maintains relative system simplicity while significantly improving analysis accuracy and error detection capability, as the system serves itself by organizing and analyzing reports through automated relevance-based clustering.
Data Source
AI summary
A computer-implemented method comprising: receiving a first report related to an application configured to run on one or more computing devices; identifying one or more terms included in the first report; identifying one or more second reports including at least one of the one or more terms; retrieving a term relevance value for a term included in at least one of the one or more second reports; determining that the term relevance value is less than a term relevance threshold value; identifying at least one of the one or more second reports for a clustering process, wherein at least one of the one or more second reports that include the term is excluded from the clustering process; implementing the clustering process using the identified at least one of the one or more second reports; and assigning the first report to a cluster.