Keyword Trend Analysis via Document Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for time-series analysis of keywords in document data are inefficient, requiring a long time to create indices and failing to provide quick responses, which is necessary for interactive operations such as web searches where results are expected within seconds.
Innovation Solution
The method involves segmenting document data by clustering or classifying based on keyword frequencies, creating document and keyword clusters, and performing time-series analysis on these clusters to acquire frequency distributions, allowing for efficient identification of trends in keyword occurrences over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If text analysis is performed on individual keywords to create indices, then measurement precision of keyword trends is improved, but loss of time increases significantly (one month required)
Solution Approach 1:
The patent segments the large set of keywords into multiple clusters based on semantic relationships and co-occurrence patterns. Instead of analyzing all keywords individually, the system creates cluster representatives that summarize groups of related keywords. This segmentation reduces the computational burden from analyzing thousands of individual keywords to analyzing a much smaller number of cluster representatives, thereby dramatically reducing index creation time while preserving trend analysis precision through the use of cluster-level aggregations.
Solution Approach 2:
The patent introduces cluster representatives as intermediary entities between individual keywords and the final trend analysis results. These representatives act as mediators that aggregate information from multiple keywords, allowing the system to capture overall trend patterns without processing every individual keyword. The cluster representatives serve as a compressed representation that maintains essential trend information while reducing computational complexity.
2Reliability
If comprehensive text analysis is performed on all document data, then reliability of trend detection is improved, but productivity decreases due to lengthy processing time
Solution Approach 1:
The patent divides the document data processing into segments organized by keyword clusters. Each cluster is processed independently to create cluster-level statistics, which are then aggregated to produce overall trend results. This segmented approach maintains reliable trend detection by preserving the relationships between related keywords while significantly improving productivity by parallelizing processing and reducing the total number of operations required.
Solution Approach 2:
The patent merges multiple individual keyword analyses into cluster-level analyses by combining keywords with similar meanings or co-occurrence patterns. This merging reduces the total number of analysis operations while maintaining trend detection reliability through the aggregation of related terms. The cluster-level merging allows the system to capture comprehensive trends more efficiently than individual keyword analysis.
3Measurement precision
If individual keyword frequencies are analyzed in detail, then measurement precision of search trends is improved, but device complexity increases due to handling large volumes of data
Solution Approach 1:
The patent segments the vast keyword space into manageable clusters, reducing device complexity by organizing data into hierarchical structures. Individual keywords are grouped into clusters based on semantic relationships, and analysis is performed at the cluster level rather than on individual keywords. This segmentation maintains measurement precision by preserving the relationships between related keywords while dramatically simplifying the data processing architecture.
Solution Approach 2:
The patent transforms the analysis from a one-dimensional individual keyword level to a two-dimensional structure with keywords organized into clusters. This dimensional change allows the system to maintain detailed keyword information while adding a cluster organization layer that reduces processing complexity. The cluster dimension provides a higher-level abstraction that simplifies data handling while preserving the ability to measure individual keyword trends when needed.
Data Source
AI summary
Processing for a time-series analysis of keywords comprises clustering or classifying pieces of document data, each of which is description of a phenomenon in a natural language, on the basis of frequencies of occurrence of keywords in the pieces of document data, individual keywords being also clustered or classified by clustering or classifying the pieces of document data, and performing a time-series analysis of frequencies of occurrence of pieces of document data containing individual keywords in clusters or classes into which the pieces of document data are clustered or classified or a time-series analysis of frequencies of occurrence of pieces of document data containing clusters or classes into which the individual keywords are clustered or classified. Frequency distribution showing variation of the frequencies of occurrence of the pieces of document data is acquired by the time-series analysis.