Digital Commons Network^™

Culture Clubs: Processing Speech By Deriving And Exploiting Linguistic Subcultures, David Guy Brizan

Dissertations, Theses, and Capstone Projects

Spoken language understanding systems are error-prone for several reasons, including individual speech variability. This is manifested in many ways, among which are differences in pronunciation, lexical inventory, grammar and disfluencies. There is, however, a lot of evidence pointing to stable language usage within subgroups of a language population. We call these subgroups linguistic subcultures.

The two broad problems are defined and a survey of the work in this space is performed. The two broad problems are: linguistic subculture detection, commonly performed via Language Identification, Accent Identification or Dialect Identification approaches; and speech and language processing tasks taken which may see …

Artificial Intelligence Hits The Barrier Of Meaning, Melanie Mitchell

Computer Science Faculty Publications and Presentations

Today’s AI systems sorely lack the essence of human intelligence: Understanding the situations we experience, being able to grasp their meaning. The lack of humanlike understanding in machines is underscored by recent studies demonstrating lack of robustness of state-of-the-art deep-learning systems. Deeper networks and larger datasets alone are not likely to unlock AI’s “barrier of meaning”; instead the field will need to embrace its original roots as an interdisciplinary science of intelligence.

A Randomized Controlled Trial On The Impact Of Polyglot Programming In A Database Context, Andreas Stefik, Phillip Merlin Uesbeck

Computer Science Faculty Research

Using more than one programming language in the same project is common practice. Often, additional languages might be introduced to projects to solve specific issues. While the practice is common, it is unclear whether it has an impact on developer productivity. In this paper, we present a pilot study investigating what happens when programmers switch between programming languages. The experiment is a repeated measures double-blind randomized controlled trial with 3 groups with various kinds of code switching in a database context. Results provide a rigorous testing methodology that can be replicated by us or others and a theoretical backing for …

Visual Entailment: A Novel Task For Fine-Grained Image Understanding, Ning Xie, Farley Lai, Derek Doran, Asim Kadav

Computer Science and Engineering Faculty Publications

Existing visual reasoning datasets such as Visual Question Answering (VQA), often suffer from biases conditioned on the question, image or answer distributions. The recently proposed CLEVR dataset addresses these limitations and requires fine-grained reasoning but the dataset is synthetic and consists of similar objects and sentence structures across the dataset. In this paper, we introduce a new inference task, Visual Entailment (VE) - consisting of image-sentence pairs whereby a premise is defined by an image, rather than a natural language sentence as in traditional Textual Entailment tasks. The goal of a trained VE model …

Application Of Virtual Reality In Rehabilitation Of Special Populations, Tingting Liu, Liu Zhen, Ping'an Qian, Rongrong Xuan, Wang Jin, Yanjie Chai

Journal of System Simulation

Abstract: With the continuous development of urbanization and the trend of population aging, the number of people who need physical and mental rehabilitation continues to increase. Using information technology to assist the rehabilitation process has the important practical significance. The existing studies about the application of virtual reality (VR) technology in the rehabilitation, psychological problems and children with autism for stroke are reviewed. A framework for the intelligent rehabilitation system is proposed. Virtual agent, the technology of somatosensory and voice recognition are advised to be used. Based on the proposed framework, a prototype rehabilitation game for autism and an upper …

Simulation Of Cyber-Physical Systems Of Systems:Some Research Areas -Computational Understanding, Awareness, And Wisdom, Tuncer Ören

Journal of System Simulation

Abstract: After a brief emphasis about the interconnected world, including Cyber-Physical Systems of Systems, the increasing importance of the decision-making by autonomous, quasi-autonomous, and autonomic systems is emphasised. Promising roles of computational understanding, computational awareness, and computational wisdom for better autonomous decision-making are outlined. The contributions of simulation-based approaches are listed.

Analyzing Public View Towards Vaccination Using Twitter, Mahajan Rutuja

Browse all Theses and Dissertations

Educating people about vaccination tends to target vaccine acceptance and reduction of hesitancy. Social media provides a promising platform for studying public perception regarding vaccination. In this study, we harvested tweets over a year related to vaccines from February 2018 to January 2019. We present a two-stage classifier to: (1) classify the tweets as relevant or non-relevant and (2) categorize them in terms of pro-vaccination, anti-vaccination, or neutral outlook. We found that the classifier was able to distinguish clearly between anti-vaccination and pro-vaccination tweets, but also misclassified many of these as neutral. Using Latent Dirichlet Allocation, we found that two …

An Evaluation Of Learning Employing Natural Language Processing And Cognitive Load Assessment, Mrunal Tipari

Dissertations

One of the key goals of Pedagogy is to assess learning. Various paradigms exist and one of this is Cognitivism. It essentially sees a human learner as an information processor and the mind as a black box with limited capacity that should be understood and studied. With respect to this, an approach is to employ the construct of cognitive load to assess a learner's experience and in turn design instructions better aligned to the human mind. However, cognitive load assessment is not an easy activity, especially in a traditional classroom setting. This research proposes a novel method for evaluating learning …

Knowledge-Enabled Entity Extraction, Hussein S. Al-Olimat

Browse all Theses and Dissertations

Information Extraction (IE) techniques are developed to extract entities, relationships, and other detailed information from unstructured text. The majority of the methods in the literature focus on designing supervised machine learning techniques, which are not very practical due to the high cost of obtaining annotations and the difficulty in creating high quality (in terms of reliability and coverage) gold standard. Therefore, semi-supervised and distantly-supervised techniques are getting more traction lately to overcome some of the challenges, such as bootstrapping the learning quickly. This dissertation focuses on information extraction, and in particular entities, i.e., Named Entity Recognition (NER), from multiple domains, …

Artificial Intelligence: How Knowledge Is Created, Transferred, And Used, Jörg Hellwig Phd, Sarah Huggett, Mark Siebert, Bamini Jayabalasingham Phd

Public Reports

This document summarizes Key Findings from the full report "Artificial Intelligence: how knowledge is created, transferred, and used", available alongside other relevant material on the Elsevier Artificial Intelligence resource centre.The RELX group has extensive data assets, powerful computing capabilities, and a vast technological talent base. These allow Elsevier to provide unique insights on AI through this report. We hope these will be of interest to research evaluators, research funders, policy makers, and researchers, as they seek to navigate this complex, evolving, and fast-growing field.

Visual Entailment Task For Visually-Grounded Language Learning, Ning Xie, Farley Lai, Derek Doran, Asim Kadav

Computer Science and Engineering Faculty Publications

We introduce a new inference task - Visual Entailment (VE) - which differs from traditional Textual Entailment (TE) tasks whereby a premise is defined by an image, rather than a natural language sentence as in TE tasks. A novel dataset SNLI-VE (publicly available at https://github.com/necla-ml/SNLI-VE) is proposed for VE tasks based on the Stanford Natural Language Inference corpus and Flickr30k. We introduce a differentiable architecture called the Explainable Visual Entailment model (EVE) to tackle the VE problem. EVE and several other state-of-the-art visual question answering (VQA) based models are evaluated on the SNLI-VE dataset, facilitating grounded language understanding and providing …

Predicting Human Interpretations Of Affect And Valence In A Social Robot, David Mcneill, Casey Kennington

Computer Science Faculty Publications and Presentations

In this paper we seek to understand how people interpret a social robot’s performance of an emotion, what we term ‘affective display,’ and the positive or negative valence of that affect. To this end, we tasked annotators with observing the Anki Cozmo robot perform its over 900 pre-scripted behaviors and labeling those behaviors with 16 possible affective display labels (e.g., interest, boredom, disgust, etc.). In our first experiment, we trained a neural network to predict annotated labels given multimodal information about the robot’s movement, face, and audio. The results suggest that pairing affects to predict the valence between them is …

Monitoring Social Media Using Machine Learning, Joseph Jinn, Keith Vanderlinden

Summer Research

Our research is an extension of prior work by CSIRO - Commonwealth Scientific and Industrial Research Organization, Australia’s national research laboratory. Our focus is on utilizing Twitter data, Tweets, as a dataset by which we measure the SLO - Social License to Operate - of various mining, gas, and oil companies. SLO is defined as the acceptability of a company’s business operations by its employees, stakeholders, and the general public. The primary purpose of the summer 2019 research project is to investigate and find a methodology by which we can effectively model the topics of all the Tweets in our …

Interpretable Distance Metric Learning For Handwritten Chinese Character Recognition, Boxiang Dong, Aparna S. Varde, Danilo Stevanovic, Jiayin Wang, Liang Zhao

Department of Computer Science Faculty Scholarship and Creative Works

Handwriting recognition is of crucial importance to both Human Computer Interaction (HCI) and paperwork digitization. In the general field of Optical Character Recognition (OCR), handwritten Chinese character recognition faces tremendous challenges due to the enormously large character sets and the amazing diversity of writing styles. Learning an appropriate distance metric to measure the difference between data inputs is the foundation of accurate handwritten character recognition. Existing distance metric learning approaches either produce unacceptable error rates, or provide little interpretability in the results. In this paper, we propose an interpretable distance metric learning approach for handwritten Chinese character recognition. The learned …

On The Mental Workload Assessment Of Uplift Mapping Representations In Linked Data, Ademar Crotti Junior, Christophe Debruyne, Luca Longo, Declan O'Sullivan

Conference papers

Self-reporting procedures have been largely employed in literature to measure the mental workload experienced by users when executing a specific task. This research proposes the adoption of these mental workload assessment techniques to the task of creating uplift mappings in Linked Data. A user study has been performed to compare the mental workload of “manually” creating such mappings, using a formal mapping language and a text editor, to the use of a visual representation, based on the block metaphor, that generate these mappings. Two subjective mental workload instruments, namely the NASA Task Load Index and the Workload Profile, were applied …

The Use Of Deep Learning Distributed Representations In The Identification Of Abusive Text, Susan Mckeever, Hao Chen, Sarah Jane Delany

Conference papers

The selection of optimal feature representations is a critical step in the use of machine learning in text classification. Traditional features (e.g. bag of words and n-grams) have dominated for decades, but in the past five years, the use of learned distributed representations has become increasingly common. In this paper, we summarise and present a categorisation of the stateof-the-art distributed representation techniques, including word and sentence embedding models. We carry out an empirical analysis of the performance of the various feature representations using the scenario of detecting abusive comments. We compare classification accuracies across a range of off-the-shelf embedding models …

A Psychosocial Behavioral Attribution Model: Examining The Relationship Between The “Dark Triad” And Cyber-Criminal Behaviors Impacting Social Networking Sites, Kim Withers

CCAC Theses and Dissertations

This study proposes that individual personality characteristics and behavioral triggering effects come together to motivate online victimization. It draws from psychology’s current understanding of personality traits, attribution theory, and criminological research. This study combines the current computer deviancy and hacker taxonomies with that of the Dark Triad model of personality mapping. Each computer deviant behavior is identified by its distinct dimensions of cyber-criminal behavior (e.g., unethical hacking, cyberbullying, cyberstalking, and identity theft) and analyzed against the Dark Triad personality factors (i.e., narcissism, Machiavellianism, and psychopathy). The goal of this study is to explore whether there are significant relationships among the …

A Comprehensive Cybersecurity Defense Framework For Large Organizations, Willarvis Smith

CCAC Theses and Dissertations

There is a growing need to understand and identify overarching organizational requirements for cybersecurity defense in large organizations. Applying proper cybersecurity defense will ensure that the right capabilities are fielded at the right locations to safeguard critical assets while minimizing duplication of effort and taking advantage of efficiencies. Exercising cybersecurity defense without an understanding of comprehensive foundational requirements instills an ad hoc and in many cases conservative approach to network security. Organizations must be synchronized across federal and civil agencies to achieve adequate cybersecurity defense. Understanding what constitutes comprehensive cybersecurity defense will ensure organizations are better protected and more efficient. …

Law's Halo And The Moral Machine, Bert I. Huang

Faculty Scholarship

How will we assess the morality of decisions made by artificial intelli­gence – and will our judgments be swayed by what the law says? Focusing on a moral dilemma in which a driverless car chooses to sacrifice its passenger to save more people, this study offers evidence that our moral intuitions can be influenced by the presence of the law.

Simulating And Modelling Opinion Dynamics, Jennifer Heermance

Graduate Research Theses & Dissertations

The foundation of social media is conversation. Social media allows people to share ideas and opinions, as well as discuss those opinions. A point of intrigue for many social scientists is how those opinions change through interaction with others. What influences someone’s opinion? When is a person willing to adapt their opinion, and when does it remain the same? Is it possible to measure these opinion dynamics? Our overall goal is to develop a more comprehensive model for opinion dynamics. The first step of this process is to simulate data that can then be analyzed and used to develop a …

Using Machine Learning Models To Discover Promising Research, Akhil Pandey Akella

Graduate Research Theses & Dissertations

Endeavors to identify valuable research involve the factors of discovery, comprehensibility, and reproducibility. The purpose of this study is to assist scholars in finding research that is both promising and of high quality. I explain how we can approach the problem of reproducibility in relation to scholarly articles and propose gauging the public understanding of science as a way to determine the comprehensibility of given research articles. Additionally, I explain how the concept of long-term social media impact supports the discovery of scholarly articles likely to be impactful even with the passage of time. I build and describe machine-learning models …

Using Reinforcement Learning In A Simulated Intelligent Tutoring System, Manohar Sai Jasti

Graduate Research Theses & Dissertations

I used reinforcement learning to investigate which categories of hints are most efficient in an intelligent tutoring system for human anatomy. Efficiency is defined as minimizing the time it takes the student to learn the material. When a student gives a wrong answer, the tutor can give them a text hint, a diagrammatic hint, or a video clip. Each type of hint takes a different amount of time to deliver and takes the student a different amount of time to understand.

I built a simulator for the intelligent tutoring system to collect data from simulated students. I implemented reinforcement learning, …

Leveraging Nlp And Social Network Analytic Techniques To Detect Censored Keywords: System Design And Experiments, Christopher S. Leberknight, Anna Feldman

Department of Computer Science Faculty Scholarship and Creative Works

Internet regulation in the form of online censorship and Internet shutdowns have been increasing over recent years. This paper presents a natural language processing (NLP) application for performing cross country probing that conceals the exact location of the originating request. A detailed discussion of the application aims to stimulate further investigation into new methods for measuring and quantifying Internet censorship practices around the world. In addition, results from two experiments involving search engine queries of banned keywords demonstrates censorship practices vary across different search engines. These results suggest opportunities for developing circumvention technologies that enable open and free access to …

Analysing The Impact Of Machine Learning To Model Subjective Mental Workload: A Case Study In Third-Level Education, Karim Moustafa, Luca Longo

Conference papers

Mental workload measurement is a complex multidisciplinary research area that includes both the theoretical and practical development of models. These models are aimed at aggregating those factors, believed to shape mental workload, and their interaction, for the purpose of human performance prediction. In the literature, models are mainly theory-driven: their distinct development has been influenced by the beliefs and intuitions of individual scholars in the disciplines of Psychology and Human Factors. This work presents a novel research that aims at reversing this tendency. Specifically, it employs a selection of learning techniques, borrowed from machine learning, to induce models of mental …

Fuzzy-Logic-Based Robust Speed Control Of Switched Reluctance Motor For Low And High Speeds, Zeki̇ Omaç

Turkish Journal of Electrical Engineering and Computer Sciences

Switched reluctance motor (SRM) is operated at high magnetic saturation to generate large torque. The flux linkage of SRM is a nonlinear function with phase current and rotor position because of the high magnetic saturation. Also, the performance of the speed controller for the SRM driver system can be negatively affected by noise, disturbances, and inertia of load torque. Therefore, the fuzzy speed controller for the SRM driver system was developed in this study. In addition, a dynamic model of SRM was simulated in Matlab/Simulink software. Based on the results obtained in this study, the speed of the SRM was …

Methods For Classification Of Consumer Review Into The Ones Written Before Or After The Product Purchase, Md Mehedi Hasan

Wayne State University Theses

Online consumer reviews provide a wealth of information about products and services that, if properly identified and extracted, could be of immense value to businesses. While classification of reviews according to sentiment polarity has been extensively studied in previous work, many more focused types of review analysis remain open problems. In this work, we introduce a novel text classification problem of separating post-purchase from pre-purchase consumer review fragments that can facilitate identification of immediate actionable insights based on the feedback from the customers, who actually purchased and own a product. To address this problem, we propose the features, which are …

The Impact Of Cross-References On The Readability Of The U.S. Internal Revenue Code, Jeffrey A. Lasky

CCAC Theses and Dissertations

Scholars and practitioners have long argued that U.S. income tax law (“the Tax Code”) is excessively complex and difficult to understand, and hence imposes non-trivial adjudication, administration, planning, and compliance costs across the spectrum of income tax stakeholders: the courts, the Internal Revenue Service, tax practitioners, business managers, and individual taxpayers. Hence, there is considerable interest in reducing the effort needed to accurately understand and apply the provisions of income tax law. Prior scholarly work has strongly argued that exceptions to Tax Code provisions as expressed by cross-references embedded in the Tax Code text constitute a major source of reading …

Efficient Algorithms For Mining Healthcare Data :, Yan Hu

Legacy Theses & Dissertations (2009 - 2024)

Data-Driven Healthcare (DDH) is defined as the usage of available medical big data to provide the best and most personalized care, which is believed to be one of the most promising directions for transforming healthcare. The healthcare data includes claims and cost data, clinical data, pharmaceutical R&D data, patient behavior and sentiment data, and health data on the web. There has been a remarkable upsurge in the adoption of healthcare data over the past several years. In particular, it has been used for medical concept extraction, patient trajectory modeling, disease inference, etc.

Emotion Forecasting In Dyadic Conversation : Characterizing And Predicting Future Emotion With Audio-Visual Information Using Deep Learning, Sadat Shahriar

Legacy Theses & Dissertations (2009 - 2024)

Emotion forecasting is the task of predicting the future emotion of a speaker, i.e., the emotion label of the future speaking turn–based on the speaker’s past and current audio-visual cues. Emotion forecasting systems require new problem formulations that differ from traditional emotion recognition systems. In this thesis, we first explore two types of forecasting windows(i.e., analysis windows for which the speaker’s emotion is being forecasted): utterance forecasting and time forecasting. Utterance forecasting is based on speaking turns and forecasts what the speaker’s emotion will be after one, two, or three speaking turns. Time forecasting forecasts what the speaker’s emotion will …

Prioritizing Interdependent Software Requirements Using Tensor And Fuzzy Graphs, Negin Misaghian, Homayun Motameni, Mohsen Rabbani

Turkish Journal of Electrical Engineering and Computer Sciences

Owing to the special stance of prioritizing tasks in requirements engineering processes, and as the requirements are not independent in nature, considering their dependencies is essential during the prioritizing process. Although different classifications of dependency types among requirements exist, only a few approaches in the prioritization process consider such valuable data (dependency among equirements). To achieve a practical prioritization, this study proposes a method based on the effects of the requirement dependencies (increase/decrease cost of) on the value of prioritization provided by the tensor concept. Since the strengths of dependencies are also influential factors in the act of prioritization, The …