Computer Science Masters Papers

Permanent URI for this collectionhdl:10365/32552

Browse

Now showing 1 - 20 of 256

Semantic enrichment of database columns using generative language models for advanced query searches
(North Dakota State University, 2024) Miska, Dylan
This study introduces a novel application of natural language generation (NLG) models to improve database table retrieval. Unlike previous works primarily utilizing embeddings and natural language processing (NLP) models, this work explores using NLGs to generate database column descriptions to enhance search accuracy. The evaluation involves two main aspects: firstly, assessing the accuracy of AI-generated column descriptions compared to ground truth descriptions; secondly, examining the impact of these descriptions when integrated into existing search models to evaluate accuracy improvements. Results indicate improved semantic alignment when comparing generated descriptions to ground truth over column names alone and improved scores for established work.
Task-Relevant API Development for Higher Education Using GraphQL
(North Dakota State University, 2024) Nygard, Dan
To develop applications that support a variety of campus needs, North Dakota State University's Enterprise Application Development team requires a method of accessing North Dakota State University System data related to university students, faculty, and staff. As state requirements limit direct access to this data, and conventional API access methods are not well-suited to application use cases, this paper will explore how the data is acquired, stored, and then made accessible to individual applications using GraphQL. A single application, Graduate Waiver Wire, is presented as a use case depicting how GraphQL aids in the automatic data update process, freeing the time previously spent by Graduate School personnel in manually updating graduate student information.
N-gram-based Search Procedure
(North Dakota State University, 2009) Woznica, Szymon
Efficient querying and discovery of meaningful patterns in data becomes more and more important with accelerating growth of data published every day on the Internet. Tree pruning-based algorithms used in most popular search programs have troubles when dealing with infrequent query strings, limiting the number of returned results that might be of interest to the user. Furthermore, the existing tools are not capable of finding data patterns that could inform the user about the frequency of occurrence and location of a specific set of words in large, user-defined sets of textual data, in an efficient manner. In this paper, we present a new search tool, which is based on n-grams and modern software technologies. Our tool can efficiently index word n-grams existing in large sets of user-defined, textual data and subsequently assist users in querying the text corpus, helping them to find hidden patterns and their locations in the input data, effectively. We describe an algorithm for extracting word n-grams with a parameter "n" equal to two, three and four, and demonstrate how it can be leveraged by the end-user of the search tool to mine data in a new way. The presented tool offers a unique feature that allows the user to search a set of n-grams, extracted from abstracts of biomedical publications obtained from the U.S. National Library of Medicine (NLM), filtering the search result by words existing in the English language. The data tier of the search tool is based on the Microsoft SQL Server 2008 supported by a set of Common Language Runtime (CLR) functions and Transact Structured Query Language (T-SQL) based stored procedures, whereas the business logic and the user interface utilizes C# .NET 3.5 libraries to support regular expression patterns, database connection (LINQ to SQL) and multithreaded system operations.
A Graphical Tool for Test Generation from State Models
(North Dakota State University, 2009) Wu, Qjpeng
This paper presented a graphical tool for generating test cases based on state models. The tool provides users with a user-friendly model editor to create their state model in a tree view structure. The tree based state model can then be saved to a disk in the form of an xml file and any existing model file can be loaded back into the tool. Two different traverse algorithms are explored by this tool, state based coverage and transition based coverage. The tool implements both algorithms and is capable of generating test paths based on different traversal algorithms. The tool also provides a code generation process that walks through these test paths and generates test cases in any one of the supported .NET based programming languages specified by the user. Lastly, the tool can generate a Visual Studio compatible model file based on the same state model created by the user. This model serves as a good visual representation of the state model created by the user in the model editor. The same state model is represented in three different forms, tree based state model in model editor, xml based state model in an xml file and graphical based state model in Visual Studio. An example is used to demonstrate the usage of this tool and the algorithms used behind the scene.
Mobile Sensor movement using Simulated Annealing
(North Dakota State University, 2009) Yamparala, Sri Harsha
Movement of mobile sensor nodes is an important aspect in wireless sensor networks, and there is considerable research underway into the design of mobile sensor movement algorithms for wireless sensor networks. Various mobile sensor movement algorithms have been proposed using different combinatorial optimization techniques such as Tahu Search, Ant Colony Optimization, and Genetic Algorithms. In this paper we present a mobile sensor movement Algorithm for Heterogenenous Sensor Networks using a Simulated annealing search technique. This paper also discusses the various factors that are taken into account while making a decision about the movement of a mobile sensor node in the event of a sensor node failure. With the help of Simulated annealing, a Movement Algorithm is proposed with an Annealing schedule that is suitable for Heterogeneous Sensor Networks, and its performance is evaluated for various sensor deployment scenarios in Heterogenenous Sensor Networks.
Design and Implementation of Social Networking Website Features
(North Dakota State University, 2010) Aakula, Srikanth Goud
People want to communicate and interact with each other over the internet. There are various websites which are helpful for this kind of socializing. The popularity and adaptability of each website depends on the features it provides to its users. This paper deals with some of features which help users extend their social-networking experience. A social-networking website is designed and developed with features targeting different areas of the social-networking environment like networking, technical support, and profile. This paper focuses on developing features which shall extend the social-networking experience of the user rather than building an entire social-networking website. The website that is developed shows the workings of the above-mentioned features in a socialnetworking website environment. The result of the paper is a working website with the features mentioned and a survey to get feedback from users. The users' satisfaction level with the features implemented can be obtained from the survey.
Neutralization of Conflict Areas Using an Ant Colony Heuristic Approach
(North Dakota State University, 2010) Bapanpally, Pavan Kumar
In this paper, we present a unique approach to solve the Neutralization of Conflict Areas problem using an Ant Colony Optimization technique. The Neutralization of Conflict Areas is a known problem, and over the years, considerable research has been conducted to find strategies [7] to solve the problem. To effectively deal with this kind of problem, many search algorithms and techniques have been proposed. The Ant Colony Optimization technique has been very successful in solving problems such as the travelling salesman problem, vehicle routing problem, and routing and scheduling problems. The suggested approach to the problem presented here is to find routes for conflict areas and then neutralize the conflict areas. To find routes to conflict areas, we used the concept of a Dynamic Source Routing protocol. To find the shortest paths, we used the Ant Colony Optimization technique. The goal of this paper is to suggest a swarm-based approach to find conflict areas, neutralize conflict areas, and recruit help from other resource units when needed to neutralize conflict areas. The proposed solution has been implemented in a simulator. We simulate how two different sets of cooperating ants called explorer ants and worker ants, with different operational abilities work together in finding and neutralizing conflict areas and also in getting help from other resources areas. The solution can be visualized using a graphical user interface. The framework that we implement will allow for experimentation with a wide variety of experimental parameters.
Teaching Encryption: A Learning Theory Approach
(North Dakota State University, 2010) Bhogadi, Manu Kishore
Bloom's taxonomy for cognitive domain is an effective taxonomy for structured learning. The six levels in Bloom's taxonomy for cognitive domain are based on the levels of difficulty. This paper focuses on teaching communication security and encryption by applying Bloom's taxonomy in the creation of educational materials and assessments. Delivering education material to the mobile phones is made possible with the advancement of mobile technologies, and the use of mobile technologies in teaching is gaining popularity because of its benefits. In this paper, delivering educational material to the mobile phones is experimented with by using a flash cards application for android based mobile phones. Experiential learning is learning by doing, is active, motivating and enables learners to retain knowledge to a greater extent. This paper provides a tool to learn RSA public key cryptographic algorithm through experiential learning.
eTablaTutor - The Electronic Tabla Leaming Game
(North Dakota State University, 2010) Bhowmick, Dibakar
Tabla is a percussion instrument widely used in Indian classical music both as a solo instrument and as an accompaniment to vocal or instrumental music. Learning Tabla, like learning any musical instrument, is very much a discontinuous process. There are periods where one makes good progress. At other times, it feels that the learner is not making any progress at all. This alternation continues indefinitely. For one who has never learned a musical instrument, the first obstacle is the most critical. It usually occurs between 1 and 3 months. It is at this point that the initial enthusiasm has worn off and there is the aching realization that learning Tabla is not going to be quick and easy. This time is when a very large number of students drop out. If one realizes ahead of time, that it is a long and interesting journey, and that it could be more interesting and fun if he just completes a few initial steps, it often reduces the dropout rate. So, I have designed this software in order to attract people's attention for learning Tabla. This software would create people's interest in playing Tabla as a game with a beautiful interface which is easy to use and learn. For the users, there is a reference manual which describes the complete details of how to use eTablaTutor. This software mixes the old music tradition of India with the new technological learning of this musical instrument.
A Clustering Approach to Improving Test Case Prioritization: An Industrial Case Study
(North Dakota State University, 2010) Carlson, Ryan Curtis
Regression testing is an important activity for controlling the quality of a software product, but it accounts for a large proportion of the costs of software. We believe that an understanding of the underlying relationships in data about software systems, including data correlations and patterns, could provide information that would help improve regression testing techniques. As an initial approach to investigating the relationships in massive data in software repositories, in this paper, we consider a clustering approach to help improve test case prioritization. We implemented new prioritization techniques that incorporate a clustering approach and utilize history data on real faults and code complexity. To assess our approach, we conducted empirical studies using an industrial software product, Microsoft Dynamics Ax, which contains real faults. Our results show that test case prioritization that utilizes a clustering approach can improve the rate of fault detection of test suites, and reduce the number of faults that slip through testing when testing activities are cut short and test cases must be omitted due to time constraints.
A Free and Flexible Poker Environment
(North Dakota State University, 2010) DeBilt, Daniel George
This paper introduces a framework and software that allows poker players to create and play original and custom poker games, through a TCP/IP connection, for free. This paper describes how the concept of playing usercreated poker games over the Internet is not known to currently exist in a flexible, private, and free environment and also critiques what currently does exist and is available for use. This paper also summarizes the software development process used and the deliverables that ultimately led to a working software application. Future version features and application extensions are also discussed that may enhance the user experience, as well as future research projects.
Alternative Clustering Algorithms in Sensor Networks
(North Dakota State University, 2010) Gupta, Divya
A wireless sensor network is composed of a large number of tiny sensor nodes that can be deployed in a variety of environments like battle fields, water, large fields, and the like, and can transmit data to a Base station (BS). In a clusterbased network organization, sensor nodes are organized into clusters and one sensor node is selected as a sensor head (SH) in each cluster. Each SH denotes a facility and sends useful information to the Base Station (BS) through other SHs via the shortest path. In this paper, we study two clustering techniques, namely kmedian clustering and k-center clustering for a wireless sensor network. All the sensor nodes are static and homogeneous (having the same specifications) and SHs are assumed to be heterogeneous with respect to other sensor nodes in their respective clusters (but homogeneous to other SHs once they are located). The focus of this paper is to compare the k-median and k-center clustering techniques based on shortest path and total intra-cluster distance. We have implemented the two clustering techniques using the Java language and necessary experimental and statistical results are provided.
Digital Deception and the Illusion of Choice: How Dark Patterns Undermine Informed Consent GDPR
(North Dakota State University, 2023) Khan, Wajeeha
Our investigation builds on prior research to examine global e-commerce data privacy, focusing on compliance with GDPR and CCPA laws introduced in 2018 and 2020. This study reveals uneven adherence to GDPR and CCPA regulations across e-commerce platforms, underscoring the persistent use of dark patterns. UK and French sites lead in GDPR compliance at 85% and 80%, while U.S. sites showed 65% adherence to CCPA. Turkish websites displayed a surprising 85% - 95% compliance with European standards. In contrast, South African platforms showed a low 30% compliance, often utilizing implicit consent methods. These findings expose significant gaps and inconsistencies in the application of data privacy laws across continents and nations. We advocate for a global standardization of data protection regulations to protect consumers and create a level playing field for businesses in the digital marketplace.
Evaluation of Convolutional Neural Networks Against Deepfakes Using Transfer Learning
(North Dakota State University, 2023) Krishan, Siddharth
The main objective of this paper is to evaluate ResNets, DenseNet, Inception and VGG, against deepfake images, to answer the question: How effectively these Convolutional Neural Network can distinguish between deepfake images and real images. The dataset was acquired from FaceForensics++ and CelebA datasets for manipulated and unmanipulated images respectively. A custom script using Python and OpenCV was applied to create the final dataset for modelling. Transfer learning is a technique of applying the learned features by a network to a similar approach. It is employed to save time and resources in training, as it does not require a large dataset to allow the network to learn effectively. The Convolutional Neural Networks are tested against different deep fakes and the networks are evaluated using metrics like precision, recall, accuracy, loss, and f-1 score. It was observed that all the networks used in the experiment performed exceptionally well, but Inception network was slightly better than the other networks in separating the real and fake images.
Sentiment Analysis of Tweets for Hate Speech Detection Using Binary Classification Algorithms and BERT
(North Dakota State University, 2023) Kaur, Manveer
In the modern world, social media wields a lot of power. Twitter, particularly, has provided people a platform to express their opinions about everything under the sun from mundane everyday life to politics, race, religion etc. It has often come under scrutiny for unabashed propagation of hate speech. This project employs natural language processing techniques on a corpus of tweets to detect hate speech. A total of 3538 unique tokens are identified that appear only in tweets classified as hate speech. With the help of data visualization techniques like word clouds and frequency distribution plots, it became evident that the occurrence of sexist, homophobic, and racist slurs is the most frequent in hate tweets. This implies that women, LGBTQ+ community, and people of color are the most targeted sections of society.
Advanced Computational Ratings for College Football Teams
(North Dakota State University, 2010) Hensley, Joel Michael
This paper explores the subject of rating systems applied to the world of college football. Current rating system methodologies are examined, and four rating systems are developed and evaluated in a program. The Hensley Rating system is introduced as a new method. The details of each of these systems are discussed, and the results are analyzed and evaluated using data from the college football seasons of 2000 - 2009.
Development Tools for Content Creation in Virtual Environments
(North Dakota State University, 2010) Hokanson, Guy Eric
lmmersive Virtual Environments (IVEs) for education are designed and implemented to enable students to learn complicated concepts in an exploratory and inquiry based manner. The environments are constructed so that multiple users can interact with the educational simulation and learn to think like a scientist. Developing these IVEs requires a multi-disciplinary development team that consists of more than just software engineers. It requires content experts to provide the information needed to create the best and most interactive lessons possible. While some content experts have a strong interest in technology and are capable of working with the development environment directly, many are more interested in their fields of expertise and would prefer to leave the programming and technical details to others. This presents a logistical problem as the experts have to somehow transfer their knowledge to the programmers who encode it into the IVE. In order to increase productivity it is suggested that web based, content editors would alleviate this development bottleneck. These tools would need to be cross platform, accessible anywhere Internet is available and not require the installation of any special software. This paper describes the design and implementation of a principled set of tools; Bot Conversation Editors used to create agent conversations, Task Editors to create and manage player tasks, and Help Editors to manage educational content in in-game reference materials.
CareCompanion: A Virtual Assistant for Enhancing Quality of Life in Alzheimer's Disease and Related Dementia Patients
(North Dakota State University, 2023) Hasan, Wordh Ul
Patients with Alzheimer's Disease and Associated Dementias (ADRD), as well as older adults, grapple with issues such as memory loss, trouble navigating, and feelings of loneliness. These challenges influence their daily routines, scheduled appointments, and interpersonal relationships. This research delves into the conceptualization, creation, and initial appraisal of "CareCompanion" – a specialized virtual aide crafted for these individuals. Utilizing state-of-the-art AI techniques like natural language processing, machine learning, and knowledge graphs, CareCompanion offers personalized reminders, guidance for navigation, and features to enhance social ties. Early assessments highlight CareCompanion's promise in elevating life quality, autonomy, and social interactions among ADRD patients and the elderly. Continued exploration and advancement promise to refine its proficiency, user-friendliness, and adaptability, meeting this group's distinct requirements and alleviating the adversities of memory gaps, navigational hurdles, and feelings of isolation.
Smart Robotic Arm Control through Autonomous Object Detection and Retrieval
(North Dakota State University, 2010) Mahodaya, Vicky Ramnath
This paper outlines the design and development of a robotic arm that can automatically detect and locate an object within its working envelope, reach to the object, grab the object with the aid of force feedback, bring it back to the user and transition back to user-control mode. A smart robotic arm like this can be used to control a brain computer interface (BCI) enabled prosthetic arm. This could be done by controlling the arm's coarse motion using BCI signals to the arm in manual mode while the fine motion control required for object capturing and retrieval could be performed by the smart robotic arm upon automatic detection of the object. The smart robotic arm was developed by interfacing force, distance and position sensors with a robotic arm possessing three actuated joints. Using feedback control and direct-inverse kinematic equations, the arm was controlled to perform autonomous object detection and retrieval.
Optimizing Incident Management Strategies Using Simulation
(North Dakota State University, 2010) Manori, Anshuman
Incidents, pre-programmed or random, are major sources of congestion on urban freeways. With many of urban freeways in the US operating close to capacity, the need to reduce the impact of incident-related congestion has become critical. Incident Management Strategies (IMS), when properly developed and deployed, have the potential to reduce such congestion on urban freeways. The purpose of this paper is to develop an analytic framework for the calibration and application of a -simulation model for testing the impact of alternate IMS on an urban transportation network. Initially a framework is presented in a conceptual form, and demonstrates the calibration and application of the model on a real life network in the Detroit metropolitan region. While the initial results are positive, full-scale validation and testing with larger networks are recommended to justify the use of -simulation techniques for assessing the impact of different IMS.

Browse

Recent Submissions