Heterogeneous Data Clustering via Input Output Space Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data clustering methods are inadequate for handling heterogeneous data types, as they often focus on homogeneous data, leading to inefficient analysis and clustering of data with mixed variable types.
Innovation Solution
A computer-automated method for clustering documents that incorporates both input and output space data, using similarity measures to aggregate documents into clusters, forming hierarchical trees, and refining clusters based on textual and numeric similarity measures to ensure homogeneity within clusters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional clustering methods are used, then clustering speed is fast, but the method cannot handle heterogeneous data types effectively
Solution Approach 1:
The patent segments the clustering process into two distinct phases: input space clustering (grouping documents by textual similarity) and output space clustering (refining groups by response type similarity). This segmentation allows the system to handle heterogeneous data types by applying appropriate similarity measures to different data segments, resolving the contradiction between versatility and complexity.
Solution Approach 2:
The patent introduces a second dimension (output space) to the traditional single-dimension clustering approach. By clustering documents first in input space and then refining in output space, the system gains the ability to handle heterogeneous data types without excessive complexity, as each dimension addresses specific aspects of the data.
2Manufacturing precision
If documents are clustered only by input space similarity, then clustering is simple, but clusters contain heterogeneous data types
Solution Approach 1:
The patent performs preliminary clustering in input space before refining in output space. This preliminary action groups documents by textual similarity first, then the output space clustering refines these groups to ensure homogeneity. This two-stage approach achieves high cluster precision without excessive complexity by building upon the preliminary grouping.
Solution Approach 2:
By adding the output space dimension to the clustering process, the patent achieves homogeneous clusters that properly separate documents with different response types. The additional dimension provides the necessary precision for handling heterogeneous data while maintaining manageable complexity through the hierarchical two-stage approach.
3Manufacturing precision
If output space similarity is computed for all clusters, then cluster homogeneity is improved, but processing time increases
Solution Approach 1:
The patent segments the clustering computation by first performing input space clustering on all documents, then applying output space similarity computation only to documents within the same input space cluster. This segmentation reduces the overall processing time while maintaining cluster homogeneity, as the computationally intensive output space comparison is limited to smaller subsets.
Solution Approach 2:
The patent performs preliminary input space clustering before computing output space similarity. This preliminary action reduces the dataset size for the subsequent output space computation, thereby reducing processing time while still achieving the desired cluster homogeneity through the refinement step.
4Ease of operation
If hierarchical tree structure is formed, then clustering organization is improved, but computational complexity increases
Solution Approach 1:
The patent segments the hierarchical tree formation into two separate trees: an input space tree for organizational structure and an output space tree for refinement. This segmentation improves clustering organization by providing clear hierarchical structure while managing computational complexity through division of the clustering task across two specialized trees.
Data Source
AI summary
A method of clustering a plurality of documents having input and output space data is disclosed that uses both input and output space criteria. The method can include aggregating documents into clusters based on input and/or output space similarity measures, and then refining the clusters based on further input and/or output space similarity measures. Aggregating the documents into clusters can include forming a hierarchical tree based on the input and/or output space similarity measures where the hierarchical tree has a root node, branching into intermediate nodes, and branching into leaf nodes covering individual documents, where the hierarchical tree includes a leaf node for each document of the plurality of documents. The method can then include forming a forest of sub-trees of the hierarchical tree based on cluster criteria. Textual and numeric similarity measures can be used depending on the type and distribution of data in the input and output spaces.