Design and Implementation of EFSC Algorithm for Hybrid Representation of Distributed Databases

In distributed database management systems, fragmenting base connections increases concurrency and hence system throughput for query processing. User queries use hybrid fragmentation methods focused on vector bindings, and deductive database implementations lack query-access-rule dependence. As a result, for hierarchical deductive information implementations, a hybrid fragmentation solution is used. The method considers the horizontal partition of base relations based on the bindings placed on user requests, then produces vertical fragments of the horizontally partitioned relations, and finally clusters rules based on attribute affinity and query and rule access frequency. The suggested fragmentation approach makes distributed deductive database structures easier to develop.

90 system (DDBS) plays a significant role (Özsu et al., 2011).DDB is a collection of logical and numerous designs that are inter-related to the computer network. The database is distributed amongst various sites and linked with multiple link speeds which frequently affect the system performance of database system, transmission cost and response time of database system (Suganya & Kalaiselvi, 2013). So, there is a need for enhancement with the design of DDBS help to improvise the data transaction availability, response time, reliability, and its performance of the system. In most of the cases, the distributed system performance is majorly affected the three design issues, namely, data allocation, fragmentation and degree of replication (Mehta et al., 2018). The term fragmentation is the procedure of portioning data into disjoint fragments, whereas the strategy for assigning fragments over multiple or single sites are referred to as data allocation. The term replication is to decide like a procedure to allocate fragments, describe the fragment and procedure for replication of fragments. However, these approaches have a significant impact on minimizing the communication cost and minimizing the response time of query amongst various sites. But, most of the application platform, the data workloads are dynamic in nature with simultaneous variation in pattern access at multiple sites. Hence, these approaches are required to understand profoundly and enhance the performance of DDBS (Nashat & Amer, 2018). Hence, the present study focuses on the optimal design of distributed database systems while consideration of data allocation and fragmentation.

Motivation of the Study
The design of DDBS is one of the most complex issues that include allocation and fragmentation (Mehta et al., 2018). The fragmentation is the distributed database system design approach toward segregate the class of database or data relation into two or more data partitions which offers database without any loss of data (Shahidul Islam Khan & Hoque, 2010). This minimizes the access of irrelevant data through database applications hence its minimize the ratio of disk accesses. The fragmentation design process is classified into three types, such as verticalfragmentation (VF), horizontal-fragmentation (HF), hybrid or mixed-fragmentation (MF) (Mehta et al., 2018). VF permits a class or relationship to be separated data into disjoint sets of attributes or columns except for the primary key. HF allows a class or relationship to be divided data into disjoint instances or tuples (Navathe et al., 1995). The term allocation is the procedure of assigning database fragments on various sites of a computer network. It may either maintain or replicated as a single copy when data is allocated.  Furthermore, we adopted association rule mining goals to discover item sets with co-occurrence of frequently performed transactions in both horizontal and vertical database process. Association rule mining process creates the relation of each item set containing each process in the needed format. For instance, if regulation is held globally and locally on data items, this specific rule was achieved data item in the real-time progression of processing item set. In addition, meta-heuristic algorithms were developed to provide near-optimal solutions to design of distributed database system. This improves the fragmentation and data allocation performance. The resulted outcome shows that we were minimized the cost function and reduces computational complexity.
During the distribution of data, it is mandatory to provide security to that data so that another party will never know their detailed information. Several approaches come for ensuring the safety to data mining is described below.
1. The data is misrepresented earlier than conveying it to the data miner.
2. The data disseminated among two or more sites, which support through a semi-honest protocol to acquire global data mining consequences without revealing any information about the data at their places.
3. Whereas by a prototypical to catalogue data, the taxonomy outcomes are only revealed to the designated party, who does not learn whatever else other than the classification results. Still, they can confirm the occurrence of specific rules lacking showing the rules.

Hybrid Fragmentation Method
In DD system, the fragmentation is used in a relational database system. But the case of the objectrelational database system, the development of the algorithm is more complex. Also, needs to consider various criteria for fragmentation in object-relational databases such as a relation among classes, attribute and sub-class usage through superclass, hierarchical structure and so on.

94
In our study, we are planning to use a hybrid  Additionally, we have selected the best execution plan to enable the query optimizer with specified weight index. By rule mining technique, the row and column were indexed from the source table.

Meta-Heuristic Method
Third, a meta-heuristic approach was applied to obtain optimized distributed deployment schemas and partition the queries into a small subset. This model hasminimized the communication cost among servers in the actual process deployment of a relational database environment. The resulted outcome helps to improve fragmentation and data allocation performance.
The implementation process for query processing using the proposed approach is, Step 1: Read and update a set of relational database queries.
Step 2: To perform separation of reading and update queries.
Step 3: Query partitions by applying hybrid fragmentation algorithm.
Step 4: query access, management and Partitions are allocated to corresponding sites by association rule mining approach Step 5: obtain optimal results by meta-heuristic

Association Rule Mining
The rule to display the employee who is eligible for getting hike based on their hire date first we are creating the data Here we have to specify two attributes.
Two major factors supports to generate the result that is support and confidence. Support is that if an employee satisfies both the rule then his/her support will be high when compared to other values. So we first initialize the support to 0 which means his/her support will be 2 which is maximum number of support. At the very first the confidence is 3 which means he /she may likely to get hike but if any of the rule failed he/she will get reduced in the confidence. In such cases the employee who is having maximum amount of support and confidence which is 2 They will be eligible to receive a full hike.
-update the commission pct value to 0.10 Displayed the employee details along with sal.

Hybrid Fragmentation
Here we are creating a fragmentation so the retrival of particular location seems faster than the traditional way of retrieving.
Contains sample output for fragmentation with exection time and cost for each opration(inside a query). For example, the timing is set for 505and the execution is processed to retrieve the result.    78-3-030-26253-2.pdf.