You are on page 1of 4

ASSIGNMENT

ALGORITHM FOR MATERIALIZE VIEW SELECTION IN


DATAWEAREHOUSE

SUBJECT: ADVANCED DATABASE MANAGEMENT SYSTEM (IT 610)

ANKUR JYOTI DAS CSI16020


PRANJAL KAKATI CSI16018
INTRODUCTION:

Data warehousing, data mining and online analytical processing represents some of the latest trends
in computing environments and information technology application to large scale processing and
analysis of data. Data warehousing technology is becoming essential for the effective business
strategy formulation and implementation. For the success of any data warehouse, accurate and
timely consolidated information along with quick and effective query response times is the basic
fundamental requirement.

Therefore, instead of computing aggregations on-the- fly every time, it is efficient if appropriate
high level views are created and saved in the database as materialized views. To answer queries, the
materialized views may be used without recomposing the views.

The materialization of all views is practically impossible because of the materialized view storage
space and maintenance cost constraint thus proper materialized views selection is one of the
intelligent decisions in designing a data warehouse to get optimal efficiency.

Materialized views required physical storage space acts just like a cache, which is copy of the data
that can be retrieved quickly. At the same time, the use of materialized views requires additional
storage space and overhead of view maintenance when refreshing the data warehouse.

WHAT IS VIEW:
A view is defined as a derived relation on top of some base relations. A view defines a function
from a set of base tables to a derived table. In data warehousing, historical data are kept in terms of
facts and dimension tables whereas aggregated values are kept in schemas like Star, Snowflake and
Fact Constellation (Chaudhuri & Dayal 1997). The derived relations from the base tables for
responding to decision support queries are called data warehouse views.

MATERIALIZE VIEW SELECTION PROBLEM:


Given a set of data warehouse queries, select a set of views to materialize so that the total query
processing cost and view maintenance cost is minimized.

This need to select an appropriate set of views to materialize for answering queries , this is denoted
as Materialize View Selection (MVS) and maintenance of the selected view denoted Maintenance of
Materialized View (MMV)

PARAMETERS CONSIDERED FOR THE ALGORITHM:


the algorithm that is discussed for materialized view selection framework is to materialize the user
views by taking into consideration of query frequency, query processing cost and storage
requirement of query. As it is impossible to materialize all the views so in a simple situation we can
compute some cost factor based on their frequency, processing cost, storage etc.

Assumptions:

QS Given set of queries


QA FQueries access frequency
T Threshold value
AL SQ Array List of selected queries
Algorithm 1:
1: begin:
2: for each query in Q S
3: find the frequency of each query Q AF
4: if ( Q AF >= T ) then
5: Add query to Array List AL SQ ;
6: end if
7: end for

Here initially set of queries are selected having access frequency greater than a threshold value T.
Next step is to compute the other factors of those selected queries.

Algorithm 2
1: begin:
3: Repeat for I 1 to Q Tot
4: Q FC Q AF /Q MFreq ;
5: Q PC Q TAE - Q TBE / Q MPT ;
6: Q SC Q RR / Q MS ;
7: Q CT Q FC ;
8: Q CT Q PC ;
9: Q CT Q SC ;
10: end repeat
Fig1
14: [Find selection cost]

Repeat for I 1 to Q Tot


S Q = * Q FC + *Q PC + (1- Q ST ) ;
Q CT S Q;
end repeat

15: [Select MV Selection Threshold]

M T = K i=K S Q / Q Tot

16: [Select materialized view having good query response, low processing and
storage cost]

Repeat for I 1 to Q Tot


SQ Q CT [i]
if (S Q >= M T ) then
Build the materialized view for the selected query
17: else
Discard the query
end repeat
REFERENCES:

[1]. Ravindra N. Jogekar, Ashish Mohod, Design and Implementation of Algorithms for
Materialized View Selection and Maintenance in Data Warehousing Environment
International Journal of Emerging Technology and Advanced Engineering, vol-3, issue 9, 2013.
[2]. Ashish Mohod, Manoj Chaudhari, Efficient Algorithms for Materialized View Selection in
Data Warehousing Environment, IJCSN , vol-2, issue 3, 2013.
[3].Rajib Goswami, D.K. Bhattacharyya, Malayananda Dutta,Approaches and Issues in View
Selection for Materializing in Data Warehouse , Department of Computer Science and
Engineering,Tezpur University, Jugal K. Kalita, Department of Computer Science, University of
Colorado Colorado Springs,

You might also like