Computer Science and Information Technology

Permanent URI for this community

https://ir.bellsuniversity.edu.ng/handle/123456789/251

Browse

Now showing 1 - 15 of 15

Metadata only
A Decision Support System For Counselling :An Analytic Hierarchy Process (AHP) Aproach.
(Bells University of Technology, 2015-09-21) OLADIBOYE, Olasunkanmi Esther; Prof. M. O. Abass
Decision Support Systems (DSS) are computer-based information systems designed to help managers to select one of t6he many alternative solutions to a problem. Put differently DSS are built to facilitate the decision-making process by process by people in institutions when the need arises. Studies have shown that there is a high rate of abuse of drug among students in tertiary institutions in Nigeria. This research is therefore an attempt to develop a decision support system (DSS) that can be useful for counselling students who have drug problem using the Analytic Hierarchical Process (AHP) model. Using the DSS developed in this study, we were able to compute Eigen vectors of the matrix of pair -wise comparison in generalized ascending order of magnitude. The highest Eigen vector was selected as the most effective method of counselling students with drug problems. The system was implemented using PHP and MySQL database. This study shows that the counselling methods such a s interview, behavior modification and counselling which refers the students to an expert, all have the same Eigen vector. That is to say that they are of equal importance with the criteria used in counselling the student. The study also revealed that 'Reward' as an approach to counselling with respect to behavioral modification has the highest priority value which is 0.6687. 'Denial' and 'Punishment' have the priority value of 0.2431 and 0.0882 respectively. In conclusion, from our findings, the best choice for counselling student with respect to behavior medication is the 'Reward Approach'.
Metadata only
A Decision Support System for Counselling: An Analytic Hierarchy Process (AHP) Aproach
(Bells University of Technology, 2015-09-30) OLADIBOYE, Olasunkanmi Esther; Prof. M. O. Abass
Decision support systems (DSS) are computer-based information systems designed to help managers to select one of the many alternative solutions to a problem put differently. DSS are built to facilitate the decision-making process by people in institution when the need arises. Studies have shown that there is high rate of abuse of drug among students in tertiary institution in Nigeria. This research is therefore an attempt to develop a decision support system (DSS) that can be useful for counselling students who have drug problem using the Analytic Hierarchical Process (AHP) model. Using the DSS developed in this study, were able to compute Eigen vectors of the matrix of pair-wise comparison in generalized ascending order of magnitude. The highest Eigen vector was selected as the most effective method of counselling students with drug problems. The system was implemented using PHP and MySQL databased. This study shows that the counselling methods such as interview, behavioral modification and counselling which refers the student to an expert, all have the same Eigen vector. That is to say that they are equal importance with the criteria used in counselling the student. The study also revealed that 'Rewards' as an approach to counselling with respect to behavioral modification has the highest priority value which is 0.6687. 'Denial' and 'Punishment' have the priority values of 0.2431 and 0.0882 respectively. In conclusion, from our findings the best choice of counselling student with respect to behavioral modification is the 'Reward Approach'.
Metadata only
Neuro-Fuzzy Approach to River Sediment Yield Prediction
(Bells University of Technology, 2016-07-28) BALOGUN, Oluwatobi; Prof. E. R. Adagunodo
The advancement of the information and communication technology has created an unlimited space for exploration in different areas in order to make meaningful observations and predict at an optimum accuracy. All around the world, data are being generated every moment in large quantities, but these data are of no use until they can be converted to useful in formation. The limitation of existing models used for predicting River sediment yield in order to forestall against natural and economic disasters forms the basis of the motivation of this research. By taking advantage of the opportunities presented by the advancement of technology, this research focuses on predicting the sediment yield from Oyan gauging station of Ogun-Osun River Basin and comparing the output with result with other existing result. Some conventional method available for sediment load/yield estimation are largely empirical. In this research, an ANFIS (Adaptive Neuro-Fuzzy inference system) is used to predict the sediment yield of Oyan gauging station of Ogun-Osun River Basin. The ANFIS is a feed forward five layered architecture and uses the back propagation algorithm to train the network. For this research, four inputs data are used, the water stage and water stage and water discharge, water temperature, and rainfall.
Metadata only
A Model For Self -Adaptive Routing Optimization in Mobile Ad-hoc Network.
(Bells University of Technology, Ota., 2017) OLUSESI, Ayobami Taiwo; Dr. S. A. Akinboro
In Mobile Ad-Hoc Network, unpredictable topology is one of the challenges that can degrade its routing performance. Particle Swarm Optimization (PSO) is a technique that has a challenge of converging prematurely when large numbers of intermediate nodes are on the network. This study designed, simulate and access the performance of Self-Adaptive Partitioned PSO (SAP-PSO) routing model in a Mobile Ad-Hoc Network (MANET). The proposed model sends information from source to destination in a 1-hop neighbor network, performed the routing optimization technique of existing PSO when numbers of intermediate nodes are few (less than one equal to ten nodes (greater than 10) are on the network, the model automatically grouped nodes into partition, and this reduces the rate of premature convergences. Input data was generated using uniformly distributed random number. The model was bench marked with the exiting POS using global best and computational time as performance metrics and simulation was carried out in MATLAB7.0 programming environment. The simulation result considered scenarios when the number of intermediate nodes is than 10 and when it is greater than 10.
Metadata only
A Data Traffic Control System (DTCS) for Optimizing Computer Network Performance.
(Bells University of technology, Ota., 2017-10-26) AKINYOKUN, Olusina Temidayo; Prof. H. O. D. Longe
The advancement of computer networks has led to the heavy utilization of network resources, which has created data traffic congestion and leads to poor network performance. An effective Data Traffic control system (DTCS) is required for the optimization of computer networks. In this research work, the developed Data Traffic Control System (DTCS) is based on network congressional control approach. Tt's an asynchronous transfer mode (ATM) congestion control system and includes a two-bucket approach to be classical leaky-bucket congestion control algorithm.
Metadata only
Privacy Enforcement on Subscriber's Data in Cloud Computing
(Bells University of Technology, Ota., 2018-04-09) ASANGA, Ukeme Joseph; Prof. O. M. Abass
Data stored in the cloud are susceptible to an array of threats from hackers and also to undetectable and unauthorized access by the cloud service provider. This is because the subscriber has no access to the internal operations of the cloud and cannot control access to his data. Therefore confidentially, availability and integrity of subscriber's data must be maintained to gain their trust in cloud-based systems. This work made use of privacy with non-trusted provider algorithm to ensure privacy of subscriber's data. Here, the subscriber encrypts their data before sending it to the se4rvfice provider that performs a second encryption before storage in the cloud. Simulation was done on multimedia dataset download from Stanford University dataset. The system results were then compared with an existing systems using a trusted cloud provided and it ensured a higher level of privacy on subscriber's data. The results were evaluated using encrypted time, decryption time and brute force hack. Results showed that with the proposed system, the subscriber had a high level of control over his data sending it to the cloud service provider. It also promises an increase adopted of cloud computing by businesses and organization with highly sensitive information as it contains measures to ensure that the subscriber's data is not readily available to unauthorized user.
Metadata only
Privacy Enforcement on Subscriber's Data in Cloud Computing
(Bells University of Technology, 2018-04-27) ASANGA, Ukeme Joseph; Prof. O. M. Abass
Data stored in the cloud are susceptible to any array of threats from hackers and also to undetectable and unauthorized access by the cloud service provider. This is because the subscriber has no access to the internal operations of the cloud and cannot control access to his data. Therefore confidentiality, availability and integrity of subscriber's data must be maintained to gain their trust in cloud-based systems. This work made use of privacy with non-trusted provider algorithm to ensure privacy of subscriber's data. here, the subscribers encrypt their data before sending it to the service provider that performs a second encryption before storage in the cloud. Simulation was done on the multimedia dataset downloaded from Stanford University dataset. The systems results were then compared with an existing system using a trusted cloud provider and it ensured a higher level of privacy on subscriber's data.
Metadata only
Twitter Sentiment Classification Using Normalized Weighted Multinomial Naive Bayes
(Bells University of Technology, 2018-08-19) IHEJIRIKA, Chiedozie Destiny; Dr. M. A. Adegoke
Twitter is a very popular micro-blogging website where individuals or group of people express their opinion and options about events, products and services leading to a pool of different contextual and unstructured messages. These messages hold different meaning or sentiments that can be useful in analysing how the public feels about certain products, events or services. Sentiment analysis involves the use of natural language processors or text analysers to extract and study subjective information in a word or group of words. The Bayesian model has been extensively used for determining sentiments in a sentence. However, the problem with the Bayesian model is its independent assumption. The independent assumption considers compound words which express a singular meaning as if each individual word of the compound word expresses a different meaning. This assumption causes a weight magnitude error during the classification and analysis of the sentiments in a message or tweet. This study addresses the weight magnitude error using the normalized weighted multinomial naïve Bayes (NWMNB). During the training of the model, the NWMNB evaluates the weight magnitude in each of the dataset, comparing and normalizing the weight (We) distribution before testing and classifying the tweet. The NWMNB was implemented and tested on benched marked data, results compared against results of other traditional naïve Bayes filters using the same dataset. The NWMNB out performs the traditional Naïve Bayes filters by 4.5 %.
Metadata only
Comparative Analysis of Certain Data Analytic Languages Using Code Based Metrics
(Bells University of Technology, 2019-10-17) ADIGUN, Maria Funmilayo; Prof. M. O. Abass
Software complexity metrics is developed and used by the various software organizations for evaluating and assuring software code quality, operation, and maintenance. Software metrics measure various types of software complexity like size metrics, control flow metrics and data flow metrics. These software complexities must be continuously calculated, followed, and controlled because high complexity may result in more errors and difficulties in maintenance, understandability, modification and testing effort. The revival of data science due to the presence of large amount of data has resulted in the need for a good programming language on which many data science applications can be developed. Hence the need to apply programming complexity metrics to existing data analytic languages so as to guide programmers in choosing tools for building data science applications. For application, binary search and quicksort algorithms are considered. The programs are written in three different data analytic languages: Python, R and Scala. Software complexity for each is found using LOC, McCabe and Halstead models. The results are compared and Scala is realized to be the most complex for all the metrics while Python and R programs are averagely at par.
Metadata only
A Mobile Application for Potholes Detection and Analysis System
(Bells University of Technology, 2020-03-08) KUDETI, Waziri Omolade; Prof. M. O. Abass
The presence of potholes on roads has over the years been a great source of concern to everyone.it has been identified as one of the major causes of road accidents as well as the wear and tear of vehicle tyres. The mobile sensing method for pothole detection is considered as the most suitable method to detect pothole for mobile devices. This is because the other methods require huge computation power for image recognition. Three different algorithms were implemented: the Z-Diff algorithm, STDEV(z)algorithm which are existing algorithms and a new improved algorithm. The data used for the experiments were obtained by driving through a selected test track which is 18.9km (11.7miles) long .it includes major single lane streets as well as minor multi-lane from Sango Ota, Ogun state to Agege, Lagos state, and is characterized by a range of degrees in road surface smoothness. This research shows that the proposed improved algorithm outperforms both the Z-DIFF algorithm and STDEV(Z) metrics. The performance metrics gives different results because each metric covers a part and considers some parameters while leaving some others. Therefore, a combination of metrics is recommended to be used to measure the performance of each algorithm against others.
Metadata only
Development of Iris Recognition System Using Enhanced Convolution Neural Network
(Bells University of Technology, Ota., 2020-09-21) ADEGBOYE, Olujoba James; Dr. Adegoke M.A.
Biometric traits such as face, iris, voice, fingerprint, and palm print have proved to be unique to each person and constant throughout its lifetime. Among all biometric characteristics, iris patterns have been revealed as one of the most reliable biometric traits to distinguish among different persons. Iris recognition is usually known as eye iris network pattern recognition technology, iris, more reliable and state for identification because iris is a unique feature which h does not change with age it remains stable and fixed from about one year of age throughout life. This research work developed an enhanced convolution neural network-based systems to recognize features. The enhancement was done using gravitational search algorithm (GSA).
Metadata only
Design of An Adaptive Intrusion Detection System Using Inspection-Based Anomaly Model
(Bells University of Technology, 2020-09-27) AJIBOYE, John; Dr. A. M. Adegoke
A common problem with analyzing intrusion detection system is the incidence of high false-positive alarm rate, which is usually caused by the in ability of the Intrusion Detection system (IDS) to accommodate changes in the pattern of users’ behaviors. This work introduces flexibility to IDS by inculcating a mechanism that enables the IDS to accommodate changes in the pattern of user’ behavior. The mechanism allows a user to authenticate his or her identity whenever the IDS observes a change in the mode of operation of IDS system. The proposed intrusion system is described as, Inspection-Based Anomaly system. The performance of the Inspection-based system was evaluated on a benched marked data. Using accuracy measure, F1, it was observed that the Inspection-based anomaly system outperforms the conventional anomaly-based system with 2.6% and outperform the best-known intrusion detection system with an accuracy of 2.8% on the same dataset. The Inspection based system is able to achieve a false positive rate that is as low as 4.9%. Therefore, the Inspection based anomaly system is recommended for use in a dynamic environment where the user’s activity is subject to changes that might occur due to unforeseen circumstance.
Metadata only
A Traffic Sign Recognition System Using Convolutional Neural Networks
(Bells University of Technology, 2020-12-10) DOSUNMU, Moyinoluwa Ayodele; Prof.H.O.D. Longe
Traffic signs are a useful invention to help keep the road safe. Although they were designed for quick and easy understanding by humans, they are not so easily recognizable by machines. A traffic sign detection recognition(TSDR) system is a technology by which a vehicle is able to recognize the traffic signs put on the road. An efficient TSDR system in which its recognition and detection mechanisms was developed based on a convolutional neural network model trained by a dataset populated with German traffic signs. The German traffic signs recognition benchmark(GTSRB) Dataset was used because it contains traffic signs that are similar to the convention adopted on Nigerian roads. five different pretrained convolutional neural networks were trained to build five models for traffic sign recognition, the model's performance was evaluated in terms of the speed at which they were trained and the accuracy at which they were able to classify the traffic signs into 43 different classes. The pretrained CNNs used are AlexNet, GoogLeNet, ResNet, VGG-16 and VGG-19 GoogLeNet performed the best with an accuracy of 95%, ResNet and VGG-19 also achieved a similar level of accuracy but they took longer period to be trained. This research shows that the accuracy of the pretrained networks at identifying traffic signs increases as the number of layers increases, also it establishes that Google Net and ResNet are the most ideal of the five pretrained CNNs experimented on, for traffic sign recognition using the GTSRB dataset
Metadata only
Digitization of Energy Production and Consumption in Beverages Industry
(Bells University of Technology, 2022-05) ALAWODE, Jeremiah Olakunle; DR.(Mrs.) Olamide O. Olusanya
Energy is a major driving force of industrial operations. Virtually nothing can be done without consumption of one or more types of energy or conversion from to another by the general law of conservation of energy. Quantifying these energies and accurately costing same has been a major concern as only a rough estimated values are usually assumed. in this research work. digitization of steam energy production and consumption in beverage industry were focused on. In the methodology adopted , flow meters identification and impartibility for energy digitization was first ascertained. Initial setup of the flow meter was carried out and each flow meter was wired using the digital input terminals selected for the pulse input signals. This is responsible for the totalizer value (initially being collected manually and locally from each flow meter) .Secondly 'the identified pulse input for each flow meter programmed in the PLC using the ladder logic programming language .This involves creating a unique tag name to identify the flow meter ,as well as programming the continuous routines responsible for totalizing the flow per pulse during each flow measurement cycle .Afterward, the connectivity configurations of each flow meter tags in the server were carried out, the OPC\OLE path is configured to ensure that data from the flow meter through the PLC gets to the digitization server. Finally, the manually obtained data from flow meter is compared with the digitized data from the server. The digital manufacturing operation (DMO) server results and the ones manually collected are compared and analyzed using the SPSS software tool for data analysis. In conclusion, the results show that 'Digitalization of Energy Production and Consumption in BEVERAGE industry is a practicable solution to solve energy measurement problem in any establishment. This solution can also be extended to other energy utilities such as water, PMS, diesel, gas, and electricity.
Metadata only
Time Series Prediction on Gated Recurrent Neural Network for Fish Mongers.
(Bells University of technology, Ota., 2023-08-23) MBAJA, Heart Uchechukwu; Dr. K. A. Sotonwa
The demand for fish and seafood products has consistently increased during the recent years since fish protein is a major animal protein consumed in many parts of the world. seafood is a very perishable product and processing ios therefore necessary to assure safety a prolonged shelf life of seafood. There are different reasons why people go into smoked fish business which can be due to, Ancestral inheritance, profit and survival.

Browse

Browsing Computer Science and Information Technology by Issue Date

Results Per Page

Sort Options