Duplicate Elimination: Difference between revisions
Jump to navigation
Jump to search
No edit summary |
No edit summary |
||
Line 42: | Line 42: | ||
[[File:Duplicate Elimination - Space.png|1000px]] | [[File:Duplicate Elimination - Space.png|1000px]] | ||
== | == Time-Space Tradeoff == | ||
[[File:Duplicate Elimination - Pareto Frontier.png|1000px]] | [[File:Duplicate Elimination - Pareto Frontier.png|1000px]] |
Revision as of 14:46, 15 February 2023
Description
SQL does not eliminate duplicates implicitly. It allows to enter duplicate values on columns other than candidate key or if did not specified any keys. If the user wants to eliminate duplicate records, he has to use DISTINCT keyword in the query.
Databases, therefore, can have duplicate entries. The problem deals with identifying and removing duplicates from a database.
Parameters
No parameters found.
Table of Algorithms
Name | Year | Time | Space | Approximation Factor | Model | Reference |
---|---|---|---|---|---|---|
Sorting based (Merge Sort) | 1964 | $O(nlogn)$ | $O(n)$ | Exact | Deterministic | |
Sorting based (Merge Sort) + real-time elimination | 1964 | $O(nlogn)$ | Exact | Deterministic | ||
BST Algorithm | 1999 | $O(nlogn)$ | $O(\log n)$ | Exact | Deterministic | |
Priority Queue Algorithm | 1976 | $O(n^{2})$ | $O(n)$ | Exact | Deterministic | |
Sorted Neighborhood Algorithm (SNA) | 1998 | $O(n^{2})$ | $O(n)$ | Exact | Deterministic | Time |
Duplicate Elimination Sorted Neighborhood Algorithm (DE-SNA) | 2002 | $O(nlogn)$ | Exact | Deterministic | ||
Adaptive Duplicate Detection Algorithm (ADD) | 2003 | $O(n^{3})$ | $O({1})$ | Exact | Deterministic | Time |