Cluster Labeling via N-gram Pair Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current clustering techniques fail to provide descriptive labels for clusters, as single words or phrases are insufficient, and automatic methods are either too lengthy or computationally intensive, lacking coverage of the cluster's content.
Innovation Solution
A system and method using n-gram selection, candidate item generation, coverage value analysis, and unique word filtration to create cluster labels, selecting pairs of n-grams that provide maximum coverage with minimal overlap, allowing for descriptive labels formed from combinations of words or phrases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If single words or phrases are used as cluster labels, then the labeling process is simple and fast, but the labels are not sufficient descriptors and fail to provide descriptive labels
Solution Approach 1:
The patent combines multiple candidate items (words or phrases) into composite labels by selecting pairs of n-grams that provide maximum coverage. This merging approach creates more descriptive labels that capture multiple themes in the cluster while maintaining efficiency through automated selection processes.
2Loss of information
If complete sentences are used as labels, then the labels provide comprehensive coverage, but they are too lengthy for many situations
Solution Approach 1:
The patent extracts key candidate items (n-grams) from the documents in the cluster and selects essential pairs to form labels. This extraction process captures the most important themes and concepts while eliminating redundant information, producing concise yet descriptive labels that avoid the length issue of complete sentences.
3Measurement precision
If complex semantic analysis is performed to create descriptive labels, then the labels are more accurate, but it is more time consuming than clustering
Solution Approach 1:
The patent uses computationally efficient n-gram frequency analysis and coverage calculations instead of complex semantic analysis. This approach uses simpler, faster operations that provide sufficient accuracy for cluster labeling without the excessive time consumption of sophisticated semantic processing.
4Productivity
If most frequent single word and/or phrase is used as label, then the labeling process is efficient, but it does not exemplify all the records in a given cluster
Solution Approach 1:
The patent merges multiple candidate items into paired labels that collectively represent diverse themes in the cluster. By selecting pairs of n-grams with maximum coverage, the system ensures that the labels exemplify a broader range of records in the cluster compared to using a single most frequent term.
Data Source
AI summary
Disclosed is a method and system for creating labels for cluster in computing environment. The system comprises receiving module, candidate items selector, combination array generator, coverage value analyzer, candidate pair selector, unique word filter and cluster label selector. Receiving module receives input data and candidate items selector selects candidate items occurring repetitively using n-gram technique to generate list of candidate items with frequency of occurrence. Combination array generator selects candidate items to populate two-dimensional array wherein each array element represents pair of n-gram. Coverage value analyzer determines coverage value for each pair of n-gram from array. Candidate pair selector selects pairs of n-gram from two-dimensional array to process and generate list of candidate pairs. The unique word filter determines number of unique words in each candidate pair. Cluster label selector sorts list of candidate pairs using coverage value and number of unique words to select cluster label.