Efficient Similarity Searching in Metric Spaces: Difference between revisions
Created page with "Title: Efficient Similarity Searching in Metric Spaces Research Question: How can we design an efficient indexing structure for similarity searching in metric spaces? Methodology: The study investigates various existing indexing structures for similarity searching in metric spaces, specifically focusing on their performance in terms of the percentage of the database scanned when varying edit distances from 10% to 100%. Results: The analysis reveals that the proposed i..." |
No edit summary |
||
| Line 9: | Line 9: | ||
Implications: The research has significant implications for applications that require similarity searching, such as multilingual databases, computational biology, text retrieval, pattern recognition, and function approximation. The proposed indexing technique can lead to substantial improvements in query processing times and overall system performance. | Implications: The research has significant implications for applications that require similarity searching, such as multilingual databases, computational biology, text retrieval, pattern recognition, and function approximation. The proposed indexing technique can lead to substantial improvements in query processing times and overall system performance. | ||
Link to Article: https://arxiv.org/abs/ | Link to Article: https://arxiv.org/abs/0403014v2 | ||
Authors: | Authors: | ||
arXiv ID: | arXiv ID: 0403014v2 | ||
[[Category:Computer Science]] | [[Category:Computer Science]] | ||
Latest revision as of 15:32, 24 December 2023
Title: Efficient Similarity Searching in Metric Spaces
Research Question: How can we design an efficient indexing structure for similarity searching in metric spaces?
Methodology: The study investigates various existing indexing structures for similarity searching in metric spaces, specifically focusing on their performance in terms of the percentage of the database scanned when varying edit distances from 10% to 100%.
Results: The analysis reveals that the proposed indexing technique, which combines clustering with M tree (MTB), outperforms the existing structures in terms of efficiency, resulting in better performance.
Implications: The research has significant implications for applications that require similarity searching, such as multilingual databases, computational biology, text retrieval, pattern recognition, and function approximation. The proposed indexing technique can lead to substantial improvements in query processing times and overall system performance.
Link to Article: https://arxiv.org/abs/0403014v2 Authors: arXiv ID: 0403014v2