Dissertations / Theses on the topic 'Computer science. Cognitive psychology'

The web has become more than a supplementary information resource but a valuable and pervasive tool for nearly all aspects of daily life including social and community participation, health promotion, creative pursuits, education, and employment opportunity. However, the web is not yet easily accessible to all people, particularly those with cognitive disabilities who encounter many challenges with access and use of the web including limited accessibility of online content and difficulty with content comprehension. Furthermore, little is documented about how individuals with cognitive disabilities who currently use the web are overcoming or being inhibited by these challenges. Much of what is documented is anecdotal or generalized as broad technical guidance rather than providing methods to empower individual end users. This research explores which websites people with cognitive disabilities use and do not use and what challenges and successes they encounter with those websites. We developed the SimpleWebAnywhere tool to address the above research needs and serve as a technology probe to determine how content simplification affects web use by people with cognitive disabilities. We explored personalizable content transformation techniques, including advertisement removal, content extraction, and text to speech, to make webpages easier to use and comprehend. We found that many people with cognitive disabilities frequently access the web for long periods of time despite popular opinion to the contrary. Web access is preferred via mobile platforms, such as smartphones and tablet computers. Users had a strong preference for entertainment content largely comprised of images, videos, and games but did not necessarily have difficulty using or understanding long, complex textual content. An intercommunity approach of combining existing open source software to provide personalized content manipulations was found to be an effective method to improve web accessibility for people with cognitive disabilities.

Authentication is the process of validating the identity of an entity, e.g., a person, a machine, etc.; the entity usually provides a proof of identity in order to be authenticated. When the entity — to be authenticated — is a human, the authentication process is called end-user authentication. Making an end-user authentication usable entails making it easy for a human to obtain, manage, and input the proof of identity in a secure manner. In machine-to-machine authentication, both ends have comparable memory and computational power to securely carry out the authentication process using cryptographic primitives and protocols. On the contrary, as a human has limited memory and computational power, in end-user authentication, cryptography is of little use. Although password based end-user authentication has many well-known security and usability problems, it is the de facto standard. Almost half a century of research effort has produced a multitude of end-user authentication methods more sophisticated than passwords; yet, none has come close to replacing passwords.

In this dissertation, taking advantage of the built-in sensing capability of smartphones, we propose an end-user authentication framework for smartphones — called ePet — which does not require any active participation from the user most of the times; thus the proposed framework is highly usable. Using data collected from subjects, we validate a part of the authentication framework for the Android platform. For web authentication, in this dissertation, we propose a novel password creation interface, which helps a user remember a newly created password with more confidence — by allowing her to perform various memory tasks built upon her new password. Declarative and motor memory help the user remember and efficiently input a password. From a within-subjects study we show that declarative memory is sufficient for passwords; motor memory mostly facilitate the input process and thus the memory tasks have been designed to help cement the declarative memory for a newly created password. This dissertation concludes with an evaluation of the increased usability of the proposed interface through a between-subjects study.

News organizations, non-profits, and even government agencies use information graphics to advertise and communicate their messages. Data visualizations are used heavily in these graphics, but they also often incorporate unusual design elements to help catch viewers’ eyes. In the struggle to rise to the top of the crowd, the data visualizations in infographics are often embellished with additions and modifications to the raw chart. The general consensus is that these embellishments can make charts less effective at communicating information, but most of them have never been tested to see if this is true. This work examines the factors in bar, pie, and donut charts that affect our perception of the charts.

I approach this in two different ways, both using a series of surveys on Mechanical Turk. The work on pie charts examines the individual contribution of arc-length, angle, and area variables so that embellishments may be evaluated based on their use of visual variables. The bar chart work examines some of the most common embellishments designers make to bar charts. This approach allows the isolated study of embellishments to determine which hinder or contribute the most to our perception of charts. I conclude with concrete recommendations based on the findings of the studies. My results show that conventional wisdom about how these charts are perceived is not always correct, and some types of embellishments are harmful while others have virtually no effect.

Visual perception depends fundamentally on statistical regularities in the environment to make sense of the world. One such regularity is the orientation anisotropy typical of natural scenes; most natural scenes contain slightly more horizontal and vertical information than oblique information. This property is likely a primary cause of the “oblique effect” in visual perception, in which subjects experience greater perceptual fluently with horizontally and vertically oriented content than oblique. However, recent changes in the visual environment, including the “carpentered” content in urban scenes and the framed, caricatured content in digital screen media presentations, may have altered the level of orientation anisotropy typical in natural scenes. Over a series of three experiments, the current work aims to evaluate whether “digital” visual experience, or visual experience with framed digital content, has the potential to alter the magnitude of the oblique effect in visual perception. Experiment 1 established a novel eye tracking method developed to index the visual oblique effect quickly and reliably using no overt responding other than eye movements. Results indicated that canonical (horizontal and vertical) contours embedded in visual noise were detected more accurately and quickly than oblique contours. For Experiment 2, the orientation anisotropy of natural, urban, and digital scenes was analyzed, in order to compare the magnitude of this anisotropic pattern across each image type. Results indicate that urban scenes contain exaggerated orientation anisotropy relative to natural scenes, and digital scenes possess this pattern to an even greater extent. Building off these two results, Experiment 3 adopts the eye tracking method of Experiment 1 as a pre- post-test measure of the oblique effect. Participants were eye tracked in the contour detection task several times before and after either a “training” session, in which they played Minecraft (Mojang, 2011) for four hours uninterrupted in a darkened room, or a “control” session, in which they simply did not interact with screens for four hours. It was predicted, based on the results of Experiment 2, that several hours of exposure to the caricatured orientation statistics of the digital stimulus would suffice to alter the magnitude of participants’ oblique effect, as indexed by the difference in the post-test relative to the pre-test. While no accuracy differences were observed in this primary manipulation, detection speed for canonical contours did alter significantly in the Minecraft subjects relative to controls. These results indicate that the oblique effect is quite robust at the level of visual contours and is measurable using eye tracking, that digital scenes contain caricatured orientation anisotropy relative to other types of scenes, and that exposure to naturalistic but caricatured scene statistics for only a few hours can alter certain aspects of visual perception.

Background: Software engineering teams must have a shared understanding of the system design in order to work independently but successfully integrate their code. These issues of understanding are important to project success but difficult to investigate with current approaches. Current techniques for investigating shared understanding, such as interviews or questionnaires, are limited by the difficulty of team members to externalize knowledge relevant to shared understanding.

Aims: This research has two goals. The first goal is to identify and validate a measure of shared understanding that researchers can use to investigate issues of shared understanding in software design. The second goal is to evaluate the potential for this measure to be used by practitioners to improve the software design process.

Method: A measure of shared understanding was developed by adapting an approach from the Team Mental Models literature. Five student teams and two industrial teams were recruited to evaluate the measure empirically. The validity of the measure, the significance of the differences in understanding found, and the applicability for design process improvement were investigated using qualitative techniques, including group interviews, observation, and questionnaires.

Results: When ranked by the measure of shared understanding, high ranking design concepts were generally, but not consistently, found to be associated with greater similarity of understanding than low ranking concepts. This supports a finding that the measure is valid, but imprecise. Although no specific misunderstandings were identified within the team, some team members found the discussion, guided by the measured differences, valuable for improving shared understanding generally.

Conclusions: The results support the use of the measure as a tool to investigate shared understanding so long as consideration is given to its limitations. It is premature for practitioners to use the measure to improve the design process. The results are based on only two industrial teams without a history of failures related to shared understanding. Future research should re-evaluate the measurement in different contexts. Guidance is given for additional research to refine the measurement.

Research on reward-driven learning has produced and substantiated theories of model-free and model-based reinforcement learning (RL), which respectively explain how humans and animals learn reflexive habits and build prospective plans. A highly developed line of work has unearthed the role of striatal dopamine in model-free learning, while the prefrontal cortex (PFC) appears to critically subserve model-based learning. The recent theory of meta-reinforcement learning (meta-RL) explained a wide array of findings by positing that the model-free dopaminergic reward prediction error trains the recurrent prefrontal network to execute arbitrary RL algorithms—including model-based RL—in its activations.

In parallel, a nascent understanding of a third reinforcement learning system is emerging: a non-parametric system that stores memory traces of individual experiences rather than aggregate statistics. Research on such episodic learning has revealed its unmistakeable traces in human behavior, developed theory to articulate algorithms underlying that behavior, and pursued the contention that the hippocampus is centrally involved. These developments lead to a set of open questions about (1) how the neural mechanisms of episodic learning relate to those underlying incremental model-free and model-based learning and (2) how the brain arbitrates among the contributions of this abundance of valuation strategies.

This thesis extends meta-RL to provide an account for episodic learning, incremental learning, and the coordination between them. In this theory of episodic meta-RL (EMRL), episodic memory reinstates activations in the prefrontal network based on contextual similarity, after passing them through a learned gating mechanism (Chapters 1 and 2). In simulation, EMRL can solve episodic contextual water maze navigation problems and episodic contextual bandit problems, including those with Omniglot class contexts and others with compositional structure (Chapter 3). Further, EMRL reproduces episodic model-based RL and its coordination with incremental model-based RL on the episodic two-step task (Vikbladh et al., 2017; Chapter 4). Chapter 5 discusses more biologically detailed extensions to EMRL, and Chapter 6 analyzes EMRL with respect to a set of recent empirical findings. Chapter 7 discusses EMRL in the context of various topics in neuroscience.

The use of language is one of the defining features of human cognition. Focusing here on two key features of language, productivity and robustness, I examine how basic questions regarding linguistic representation can be approached with the help of probabilistic generative language models, or PGLMs. These statistical models, which capture aspects of linguistic structure in terms of distributions over events, can serve as both the product of language learning and as prior knowledge in real-time language processing. In the first two chapters, I show how PGLMs can be used to make inferences about the nature of people's linguistic representations. In Chapter 1, I look at the representations of language learners, tracing the earliest evidence for a noun category in large developmental corpora. In Chapter 2, I evaluate broad-coverage language models reflecting contrasting assumptions about the information sources and abstractions used for in-context spoken word recognition in their ability to capture people's behavior in a large online game of “Telephone.” In Chapter 3, I show how these models can be used to examine the properties of lexicons. I use a measure derived from a probabilistic generative model of word structure to provide a novel interpretation of a longstanding linguistic universal, motivating it in terms of cognitive pressures that arise from communication. I conclude by considering the prospects for a unified, expectations-oriented account of language processing and first language learning.

The ability to reason analogically is a central marker of human-level cognition. Analogy involves mapping, reorganizing, and creating structural knowledge, a particular type of cognitive construct commonly understood as residing purely within the domain of declarative knowledge. Yet existing computational models of analogy struggle to show human-level performance on any data sets not manually constructed for the purposes of demonstration, a problem referred to as the tailorability concern. Solving the tailorability concern may require more investigation into the nature of cognitive structures, defined as those elements in mental representation which are referred to whenever contemporary literature on analogy discusses "structured" knowledge.

I propose to develop the theory of Analogical Constructivism. This theory builds on Piaget's constructivist epistemology, first refining its concepts by clarifying the modifications Piaget himself made in his later, less-discussed works. I reconcile Piaget's assertion that meaning is, first and foremost, rooted in the action schemas that the agent is both born with and develops throughout life, with an account of cognitive structure, concluding that cognitive structure is inseparable from action-centered/procedural knowledge.

After a defense of the claim that cognitive structure cannot exist apart from actions (a claim which I refer to as "No-semantically-empty-structure"), I introduce PAGI World, a simulation environment rich enough in possible actions to foster the growth of artificial agents capable of producing their own cognitive structures. I conclude with a brief demonstration of an agent in PAGI World, and discuss future work.

Neuromodulation is a neurophysiological process by which a single neuron can regulate the neural activity of a diverse population of neurons. Sensory integration is a neurobiological process by which the brain combines multiple sensory modality inputs (i.e., vision, proprioception, audition, tactile, olfactory, vestibular, interoception, and taste) into usable functional outputs. In biological systems, neuromodulation and sensory integration have been shown to have a strong influence over action selection (decision-making) and action execution (motor output) respectively. The experiments portrayed in Chapters 1-4 provide empirical and theoretical evidence for neuromodulatory influence over selected actions through predictions of expected costs and rewards. The simulation experiments described in Chapters 5-6 illustrate how sensory integration influences action execution across different neural architectures in visually and memory guided sensorimotor transformation tasks. The implications of these results and future endeavors are discussed in Chapter 7, along with a proposed computational model of both action selection and sensory integration to investigate the dynamics of decision-making influenced by the integration of multiple sensory inputs in order to execute an action.

We introduce a novel approach to evolving spike neural network (SNN) based Spatio-temporal (ST) pattern classifiers that can detect occurrences of hidden structures in a ST data. We test this learning paradigm to find characteristic electrical patterns in visually evoked response potentials (VERPs) generated by an alcoholic brain.

We cast the alcoholic classification task as a multiple feature selection (FS) problem. The FS problems are grouped under 2 classes: the spatial task and the temporal task. The objective of the spatial FS task is to choose a correct subset of electroencephalogram (EEG) leads (the spatial-features) along with the lead-weighs (numeric attributes) using which a composite signal can be created. The temporal FS task involves detecting temporal patterns that occur more frequently in the alcoholic composite signals than in the control signals. To facilitate the evolution of such a classifier, we introduce design rules for SNN based temporal pattern detectors (TPDs) and novel crossover operators for the simultaneous FS task.

The conventional techniques for characterizing the alcoholic VERPs use the information in the gamma-band (30 to 50 Hz) to develop a set of feature vectors and then train a classifier using these feature vectors. Using the SNN based evolutionary learning paradigm we were able to solve this problem in 1 step; the SNN performed both temporal feature extraction and classification. Unlike the conventional techniques we did not make any specific assumptions regarding the spectral characteristics of the data; we did not implement a gamma-band filter. Also, we located regions on the skull of an alcoholic subject that produced abnormal electrical activity compared to the controls. These regions are consistent with prior findings in the literature. The classification accuracy was measured as the area under the receiver operator characteristic curve (ROC). The area under the ROC curve for the training set varied from 90.32% to 98.83% and for the testing set it varied from 87.17% to 95.9%.

For every participant role filler in an utterance, speakers must choose to leave it bare (e.g., "the interviewer") or to modify it (e.g., "the interviewer on Fresh Air"). Their decision is the end result of a combination of complex factors ranging from the original message to how distracted the speaker is. When we use corpora to create language models, part of our job is understanding the observable properties in and around an event description that allow us to predict these decisions. A considerable body of work on language production and discourse pragmatics concentrates on measuring noun phrase predictability and other forms of shared knowledge that help determine the balance point between over- and under-specification of a participant role filler. Although the importance of predictability as measured by long-term probabilities has long been recognized, I present a novel quantitative analysis of participant role filler predictability, the structure of the mental lexicon, and how the interaction of these two inform a speaker's internal perception of informativity. Standard Gricean assumptions tend to be efficiency oriented. Speakers will be informative enough but not wastefully so. Using these to model corpus distributions predict that noun phrase modification rates are directly proportional to predictability in order to satisfy the speaker's obligation to always be informative. In contrast, standard Firthian models (built around the idea that "you know a word by the company it keeps") assume spreading activation—and not efficiency—is the dominant predictor of usage. Sensitivity to activation's effect predicts that noun phrase modification rates are inversely proportional to predictability. Strongly connected participant role fillers could be easily activated for production while weakly connected participant role fillers would either be mentioned less often or themselves trigger strongly connected features (not normally associated with the head verb) to be primed for production.

To distinguish between these competing assumptions, I analyze participant role filler modification rates in event descriptions with respect to three indicators: the syntactic and semantic optionality of the role filler, the general predictability of the verb's role fillers, and the predictability of individual pairs of verb/participant role fillers. First, I use insights from linguistic theory to classify verbs and their participant roles into classes of syntactic optionality and semantic optionality. Second, I quantify over a large corpus the general predictability of a verb's participant roles and the specific predictability of each pair of verb/participant role filler. Finally, I model the relationship between the three indicators and modification in order to ascertain whether speakers have a stronger tendency to modify the more predictable participant role fillers, as Grice's Maxim of Relevancy predicts, or a tendency to modify the less predictable participant role fillers, as a Firthian activation-based model predicts.

I present descriptive statistical models to chart the relationship between predictability, syntactic optionality of a participant role, and semantic optionality of a participant role. In general, verb classes with stronger mental lexicon connections to their participant role fillers according to theory also have more predictable participant role fillers in the British National Corpus. Specifically, syntactically optional direct object verbs and semantically obligatory instrument verbs have more predictable participant role fillers than the opposite, comparable verb class. I also present several linear mixed-effect models to determine how predictive of modification the independent variables of syntactic verb class, semantic verb class, and verb/participant role filler predictability are. According to these models, speakers are significantly more likely to modify the less predicted participant role fillers even when taking into account individual verb and verb class differences. I conclude that mental lexicon accessibility modulates noun phrase realization according to a Firthian activation-based model. For each factor, I discuss possible explanations for the correlations between modification, predictability, and optionality and how these correlations make sense within a larger production model.

Survey theory developed as a means to overcome problems with design and analysis is inherent in early research. Survey sampling methodology improves the quality of information collected, ensures the accuracy of data analysis, and reduces the cost of research. Technology drives the evolution of data collection and analysis that is required in survey sampling. In turn, this influences survey sampling techniques. I investigate the history of survey sampling, current survey sampling theory, and current theory applied to two examples: 1) a stratified market research survey, and 2) a psychological survey for health science research.

The market research survey was an original design using a specific methodology: conduct pre-interviews on a small sample, develop survey questions based on the qualitative research, stratify the target sample during data collection, and perform data analysis on the resulting cross-sectional data. The second survey utilizes well-developed and tested health measurement instruments that have already been developed and tested. The resulting longitudinal data are then scored and analyzed.

During an extravehicular activity (EVA), the role of an astronaut involves a multitude of complex tasks. Whether that task is a science experiment aboard the International Space Station, or traversing extraterrestrial terrain – attention, communication, and instruction are essential. As an aid, augmented reality (AR) can portray suit informatics and procedures within line-of-sight while minimizing attentional loss. Currently, there exists little research highlighting the human systems considerations to qualify AR systems for space suit applications. This study quantifies user interface (UI) and human performance measures for an AR prototype on the Mark III space suit. For user testing, 21 military pilots and personnel (11 men, 10 women) evaluated UI search tasks and completed a series of AR-instructed EVA dexterity tasks in an elevated luminosity, background clutter, and workload scenario. UI results suggest correlations for readability and usability; whereas, human performance results provide situational awareness, workload, and task performance data.

How mental representations are constructed and how they evolve are central problems for cognitive science. Representation decisions help determine what computations are hard or easy. Structured, relational representations are a hallmark of human cognition. Developmental studies show that children do not perform as well as adults in tasks that require noticing relational similarity. What drives this development? Gentner and her colleagues have argued that comparison and language are two forces driving this change. This thesis explores these ideas further by presenting a computational model of forced choice tasks to illuminate the roles of comparison and language in driving representational change.

The model simulates the following roles of comparison. First, comparison can be used to make selections in forced-choice tasks. Second, comparisons from recent positive experiences are assimilated as interim generalizations which are retrieved for subsequent tasks and influence encoding by highlighting relevant structure. Third, comparisons suggest opportunities for re-representation. Finally, verifying candidate inferences resulting from a comparison provides a way to augment encodings with background knowledge, thus enriching representations. The model simulates the role of language in facilitating the creation and enrichment of generalizations as follows. When two objects are given the same label, the model compares them. This leads to an interim generalization associated with that label, enriched with commonalities from background knowledge.

We tested these hypotheses by extending the Companion cognitive architecture and simulating three developmental studies. To reduce tailorability, the visual stimuli were provided as sketches and the objects were labeled using simplified English. The model was evaluated by comparing its behavior and learning trajectory to that of children in the developmental studies. The performance of the model in the simulations provide evidence for the claims of this thesis.

Sex differences in spatial ability, especially mental rotation, navigation and object-location memory are described in this essay. Biological differences in brain morphology, hormones and genes between men and women are presented as explanations for the sex differences. Another level of explanations offered are evolutionary, hence the most influential evolutionary psychological theories are summarized and evaluated. These theories are Gaulin’s and Fitzgerald’s male range theory, Silverman’s and Eals’s hunter-gatherer theory, and Ecuyer-Dab’s and Robert’s twofold selection theory. The hunter-gatherer theory at present seems to be of the most importance, though the twofold selection theory may in the future challenge it. Regardless, united biological and evolutionary explanations would create the best comprehensive theory.

In this work we explore how close the artificial intelligence community has come to model the human mind regarding representation and processing of sequences. We analyse results produced by cognitive psychologists, who explore real minds, for features exhibited by human short- and long-term memory when representing and processing sequences. We compare these features with theories and models from the AI community divided into two types of theories: intrinsic and extrinsic theories. We conclude that the intrinsic theories have managed to explain most of the features, whereas the extrinsic theories still have a lot to do before exhibiting all features. We also present several suggestions for continued research to the AI community within the area of sequence representation and processing in the human mind.