Hemakumar Reddy Front Pages
Hemakumar Reddy Front Pages
Submitted
By
C. HEMAKUMAR REDDY
Regd. No : 2722132
Under the valuable Guidance of
Mr. A. MURALI MOHAN KUMAR, M.C.A, HSM, GNIIT
ASSOSIATE PROFFESSOR
2022-2024
Dear C HEMAKUMAR REDDY,
We look forward to working with you and wish you success in your career with us.
Warm regards,
PENCHALA SWAMY. P
PALAMANER – 517408
2022-2024
CERTIFICATE
1.
2.
DECLARATION
I hereby declare that the project work entitled PERSONALITY - AND VALUE-AWARE
SCHEDULING OF USER REQUESTS IN CLOUD FOR PROFIT MAXIMIZATION
fulfillment of the requirements for the award of the degree of Master of Computer
Applications and it is a report of the original work done by me under the guidance of
Associate Prof. Mr. A. Murali Mohan Kumar during the period of May – July 2024.
C HEMAKUMAR REDDY
ABSTRACT
The main goal of a cloud provider is to make profits by providing services to users. Existing
profit optimization strategies employ homogeneous user models in which user personality is
ignored, resulting in fewer profits and particularly notably lower user satisfaction that in turn,
leads to fewer users and reduced profits. In this paper, we propose efficient personality-aware
request scheduling schemes to maximize the profit of the cloud provider under the constraint
of user satisfaction. Specifically, we first model the service requests at the granularity of
individual personality and propose a personalized user satisfaction prediction model based on
questionnaires. Subsequently, we design a personality-guided integer linear programming
(ILP)-based request scheduling algorithm to maximize the profit under the constraint of user
satisfaction, which is followed by an approximate but lightweight value assessment and cross
entropy (VACE)-based profit improvement scheme. The VACE-based scheme is especially
tailored for applications with high scheduling resolution. Extensive simulation results show
that our satisfaction prediction model can achieve the accuracy of up to 83%, and our profit
optimization schemes can improve the profit by at least 3.96% as compared to the
benchmarking methods while still obtaining a speedup of at least 1.68x.
PERSONALITY - AND VALUE-AWARE SCHEDULING OF
USER REQUESTS IN CLOUD FOR PROFIT MAXIMIZATION
INDEX