Data Mining and Knowledge Engineering

Pradosh Chandra Pattnaik ¹, Md. Shafakhatullah Khan ¹, V.S.V. Mallikharjuna Rao ²

Affiliations
1 Department of Information Technology and Computer Applications, Aurora's Engineering College, Jawaharlal Nehru Technological University, Hyderabad, Andhra Pradesh, IN
2 Department of Computer Science, Aurora's PG College, Osmania University, Hyderabad, Andhra Pradesh, IN

Abstract

This paper proposes a new idea which makes the processes not to wait for a long time to enter into a critical section to access a data structure which is already assigned to the other process and that process enters into critical section. Till date so many approaches and algorithms are proposed to reduce the delay time between the successive executions of the critical section, but no one proposed how to assign the data structure to all the process who are requesting for the data structure. In this paper the idea we are proposing is to create an instance of the data structure before assigning to the process and allow each and every process into critical section to avoid the delay time permanently. The proposed algorithm every time creates an instance of the data structure before assigning to the requested process and simultaneously it checks for the instance released by the process already assigned instanced data structure. If when a new process is requesting for the data structure and at the same time already assigned process releases the instance then the coordinator will do both creation and deletion of the instance.

Keywords

Mutual Exclusion, Critical Section, Resource Management, Group Mutual Exclusion.

Full Text

     

Md. Shafakhatullah Khan ¹, A. Chandrashekar Sharma ², Shaik Shah Nawaz ²

Affiliations
1 Department of Information Technology and Computer Applications, Aurora's Engineering College, Jawaharlal Nehru Technological University, Hyderabad, Andhra Pradesh, IN
2 Department of Computer Science & Engineering, Aurora's Engineering College, Jawaharlal Nehru Technological University, Hyderabad, Andhra Pradesh, IN

Abstract

Spatial data is represented with X and Y coordinates (or latitude and longitude) which might be a fixed point in a map or the position of an object in a map. In this paper we are using vector model for representing spatial data. Probabilistic spatial data is one where the coordinates or the position of an object may be one of 'n' different locations. If there are 'm' such objects then finding the probabilistic nearest neighborhood of such objects is known as spatial query. Finding the nearest neighbor (NN) of probabilistic spatial data is quite challenging because from time to time the objects will not be in the same position. In this paper we propose a fast and efficient method for finding the NN with linear optimization of position vectors. The method first finds the probable objects whose any of the position is nearer to the query object and then optimized amongst obtained result set to find the precise NN-Candidate. Result shows significant closeness of the query object with the obtained NN-candidate.

Keywords

Nearest Neighbor, Probabilistic Spatial Data, Linear Optimization.

Full Text