Theses and Dissertations
This collection includes both ASU Theses and Dissertations, submitted by graduate students, and the Barrett, Honors College theses submitted by undergraduate students.
Displaying 1 - 10 of 99
Description
The increasing popularity of Twitter renders improved trustworthiness and relevance assessment of tweets much more important for search. However, given the limitations on the size of tweets, it is hard to extract measures for ranking from the tweet's content alone. I propose a method of ranking tweets by generating a reputation score for each tweet that is based not just on content, but also additional information from the Twitter ecosystem that consists of users, tweets, and the web pages that tweets link to. This information is obtained by modeling the Twitter ecosystem as a three-layer graph. The reputation score is used to power two novel methods of ranking tweets by propagating the reputation over an agreement graph based on tweets' content similarity. Additionally, I show how the agreement graph helps counter tweet spam. An evaluation of my method on 16~million tweets from the TREC 2011 Microblog Dataset shows that it doubles the precision over baseline Twitter Search and achieves higher precision than current state of the art method. I present a detailed internal empirical evaluation of RAProp in comparison to several alternative approaches proposed by me, as well as external evaluation in comparison to the current state of the art method.
ContributorsRavikumar, Srijith (Author) / Kambhampati, Subbarao (Thesis advisor) / Davulcu, Hasan (Committee member) / Liu, Huan (Committee member) / Arizona State University (Publisher)
Created2013
Description
Automating aspects of biocuration through biomedical information extraction could significantly impact biomedical research by enabling greater biocuration throughput and improving the feasibility of a wider scope. An important step in biomedical information extraction systems is named entity recognition (NER), where mentions of entities such as proteins and diseases are located within natural-language text and their semantic type is determined. This step is critical for later tasks in an information extraction pipeline, including normalization and relationship extraction. BANNER is a benchmark biomedical NER system using linear-chain conditional random fields and the rich feature set approach. A case study with BANNER locating genes and proteins in biomedical literature is described. The first corpus for disease NER adequate for use as training data is introduced, and employed in a case study of disease NER. The first corpus locating adverse drug reactions (ADRs) in user posts to a health-related social website is also described, and a system to locate and identify ADRs in social media text is created and evaluated. The rich feature set approach to creating NER feature sets is argued to be subject to diminishing returns, implying that additional improvements may require more sophisticated methods for creating the feature set. This motivates the first application of multivariate feature selection with filters and false discovery rate analysis to biomedical NER, resulting in a feature set at least 3 orders of magnitude smaller than the set created by the rich feature set approach. Finally, two novel approaches to NER by modeling the semantics of token sequences are introduced. The first method focuses on the sequence content by using language models to determine whether a sequence resembles entries in a lexicon of entity names or text from an unlabeled corpus more closely. The second method models the distributional semantics of token sequences, determining the similarity between a potential mention and the token sequences from the training data by analyzing the contexts where each sequence appears in a large unlabeled corpus. The second method is shown to improve the performance of BANNER on multiple data sets.
ContributorsLeaman, James Robert (Author) / Gonzalez, Graciela (Thesis advisor) / Baral, Chitta (Thesis advisor) / Cohen, Kevin B (Committee member) / Liu, Huan (Committee member) / Ye, Jieping (Committee member) / Arizona State University (Publisher)
Created2013
Description
When surgical resection becomes necessary to alleviate a patient's epileptiform activity, that patient is monitored by video synchronized with electrocorticography (ECoG) to determine the type and location of seizure focus. This provides a unique opportunity for researchers to gather neurophysiological data with high temporal and spatial resolution; these data are assessed prior to surgical resection to ensure the preservation of the patient's quality of life, e.g. avoid the removal of brain tissue required for speech processing. Currently considered the "gold standard" for the mapping of cortex, electrical cortical stimulation (ECS) involves the systematic activation of pairs of electrodes to localize functionally specific brain regions. This method has distinct limitations, which often includes pain experienced by the patient. Even in the best cases, the technique suffers from subjective assessments on the parts of both patients and physicians, and high inter- and intra-observer variability. Recent advances have been made as researchers have reported the localization of language areas through several signal processing methodologies, all necessitating patient participation in a controlled experiment. The development of a quantification tool to localize speech areas in which a patient is engaged in an unconstrained interpersonal conversation would eliminate the dependence of biased patient and reviewer input, as well as unnecessary discomfort to the patient. Post-hoc ECoG data were gathered from five patients with intractable epilepsy while each was engaged in a conversation with family members or clinicians. After the data were separated into different speech conditions, the power of each was compared to baseline to determine statistically significant activated electrodes. The results of several analytical methods are presented here. The algorithms did not yield language-specific areas exclusively, as broad activation of statistically significant electrodes was apparent across cortical areas. For one patient, 15 adjacent contacts along superior temporal gyrus (STG) and posterior part of the temporal lobe were determined language-significant through a controlled experiment. The task involved a patient lying in bed listening to repeated words, and yielded statistically significant activations that aligned with those of clinical evaluation. The results of this study do not support the hypothesis that unconstrained conversation may be used to localize areas required for receptive and productive speech, yet suggests a simple listening task may be an adequate alternative to direct cortical stimulation.
ContributorsLingo VanGilder, Jennapher (Author) / Helms Tillery, Stephen I (Thesis advisor) / Wahnoun, Remy (Thesis advisor) / Buneo, Christopher (Committee member) / Arizona State University (Publisher)
Created2013
Description
Intracortical microstimulation (ICMS) within somatosensory cortex can produce artificial sensations including touch, pressure, and vibration. There is significant interest in using ICMS to provide sensory feedback for a prosthetic limb. In such a system, information recorded from sensors on the prosthetic would be translated into electrical stimulation and delivered directly to the brain, providing feedback about features of objects in contact with the prosthetic. To achieve this goal, multiple simultaneous streams of information will need to be encoded by ICMS in a manner that produces robust, reliable, and discriminable sensations. The first segment of this work focuses on the discriminability of sensations elicited by ICMS within somatosensory cortex. Stimulation on multiple single electrodes and near-simultaneous stimulation across multiple electrodes, driven by a multimodal tactile sensor, were both used in these experiments. A SynTouch BioTac sensor was moved across a flat surface in several directions, and a subset of the sensor's electrode impedance channels were used to drive multichannel ICMS in the somatosensory cortex of a non-human primate. The animal performed a behavioral task during this stimulation to indicate the discriminability of sensations evoked by the electrical stimulation. The animal's responses to ICMS were somewhat inconsistent across experimental sessions but indicated that discriminable sensations were evoked by both single and multichannel ICMS. The factors that affect the discriminability of stimulation-induced sensations are not well understood, in part because the relationship between ICMS and the neural activity it induces is poorly defined. The second component of this work was to develop computational models that describe the populations of neurons likely to be activated by ICMS. Models of several neurons were constructed, and their responses to ICMS were calculated. A three-dimensional cortical model was constructed using these cell models and used to identify the populations of neurons likely to be recruited by ICMS. Stimulation activated neurons in a sparse and discontinuous fashion; additionally, the type, number, and location of neurons likely to be activated by stimulation varied with electrode depth.
ContributorsOverstreet, Cynthia K (Author) / Helms Tillery, Stephen I (Thesis advisor) / Santos, Veronica (Committee member) / Buneo, Christopher (Committee member) / Otto, Kevin (Committee member) / Santello, Marco (Committee member) / Arizona State University (Publisher)
Created2013
Description
Humans' ability to perform fine object and tool manipulation is a defining feature of their sensorimotor repertoire. How the central nervous system builds and maintains internal representations of such skilled hand-object interactions has attracted significant attention over the past three decades. Nevertheless, two major gaps exist: a) how digit positions and forces are coordinated during natural manipulation tasks, and b) what mechanisms underlie the formation and retention of internal representations of dexterous manipulation. This dissertation addresses these two questions through five experiments that are based on novel grip devices and experimental protocols. It was found that high-level representation of manipulation tasks can be learned in an effector-independent fashion. Specifically, when challenged by trial-to-trial variability in finger positions or using digits that were not previously engaged in learning the task, subjects could adjust finger forces to compensate for this variability, thus leading to consistent task performance. The results from a follow-up experiment conducted in a virtual reality environment indicate that haptic feedback is sufficient to implement the above coordination between digit position and forces. However, it was also found that the generalizability of a learned manipulation is limited across tasks. Specifically, when subjects learned to manipulate the same object across different contexts that require different motor output, interference was found at the time of switching contexts. Data from additional studies provide evidence for parallel learning processes, which are characterized by different rates of decay and learning. These experiments have provided important insight into the neural mechanisms underlying learning and control of object manipulation. The present findings have potential biomedical applications including brain-machine interfaces, rehabilitation of hand function, and prosthetics.
ContributorsFu, Qiushi (Author) / Santello, Marco (Thesis advisor) / Helms Tillery, Stephen (Committee member) / Buneo, Christopher (Committee member) / Santos, Veronica (Committee member) / Artemiadis, Panagiotis (Committee member) / Arizona State University (Publisher)
Created2013
Description
Data mining is increasing in importance in solving a variety of industry problems. Our initiative involves the estimation of resource requirements by skill set for future projects by mining and analyzing actual resource consumption data from past projects in the semiconductor industry. To achieve this goal we face difficulties like data with relevant consumption information but stored in different format and insufficient data about project attributes to interpret consumption data. Our first goal is to clean the historical data and organize it into meaningful structures for analysis. Once the preprocessing on data is completed, different data mining techniques like clustering is applied to find projects which involve resources of similar skillsets and which involve similar complexities and size. This results in "resource utilization templates" for groups of related projects from a resource consumption perspective. Then project characteristics are identified which generate this diversity in headcounts and skillsets. These characteristics are not currently contained in the data base and are elicited from the managers of historical projects. This represents an opportunity to improve the usefulness of the data collection system for the future. The ultimate goal is to match the product technical features with the resource requirement for projects in the past as a model to forecast resource requirements by skill set for future projects. The forecasting model is developed using linear regression with cross validation of the training data as the past project execution are relatively few in number. Acceptable levels of forecast accuracy are achieved relative to human experts' results and the tool is applied to forecast some future projects' resource demand.
ContributorsBhattacharya, Indrani (Author) / Sen, Arunabha (Thesis advisor) / Kempf, Karl G. (Thesis advisor) / Liu, Huan (Committee member) / Arizona State University (Publisher)
Created2013
Description
Contemporary online social platforms present individuals with social signals in the form of news feed on their peers' activities. On networks such as Facebook, Quora, network operator decides how that information is shown to an individual. Then the user, with her own interests and resource constraints selectively acts on a subset of items presented to her. The network operator again, shows that activity to a selection of peers, and thus creating a behavioral loop. That mechanism of interaction and information flow raises some very interesting questions such as: can network operator design social signals to promote a particular activity like sustainability, public health care awareness, or to promote a specific product? The focus of my thesis is to answer that question. In this thesis, I develop a framework to personalize social signals for users to guide their activities on an online platform. As the result, we gradually nudge the activity distribution on the platform from the initial distribution p to the target distribution q. My work is particularly applicable to guiding collaborations, guiding collective actions, and online advertising. In particular, I first propose a probabilistic model on how users behave and how information flows on the platform. The main part of this thesis after that discusses the Influence Individuals through Social Signals (IISS) framework. IISS consists of four main components: (1) Learner: it learns users' interests and characteristics from their historical activities using Bayesian model, (2) Calculator: it uses gradient descent method to compute the intermediate activity distributions, (3) Selector: it selects users who can be influenced to adopt or drop specific activities, (4) Designer: it personalizes social signals for each user. I evaluate the performance of IISS framework by simulation on several network topologies such as preferential attachment, small world, and random. I show that the framework gradually nudges users' activities to approach the target distribution. I use both simulation and mathematical method to analyse convergence properties such as how fast and how close we can approach the target distribution. When the number of activities is 3, I show that for about 45% of target distributions, we can achieve KL-divergence as low as 0.05. But for some other distributions KL-divergence can be as large as 0.5.
ContributorsLe, Tien D (Author) / Sundaram, Hari (Thesis advisor) / Davulcu, Hasan (Thesis advisor) / Liu, Huan (Committee member) / Arizona State University (Publisher)
Created2014
Description
Advances in implantable MEMS technology has made possible adaptive micro-robotic implants that can track and record from single neurons in the brain. Development of autonomous neural interfaces opens up exciting possibilities of micro-robots performing standard electrophysiological techniques that would previously take researchers several hundred hours to train and achieve the desired skill level. It would result in more reliable and adaptive neural interfaces that could record optimal neural activity 24/7 with high fidelity signals, high yield and increased throughput. The main contribution here is validating adaptive strategies to overcome challenges in autonomous navigation of microelectrodes inside the brain. The following issues pose significant challenges as brain tissue is both functionally and structurally dynamic: a) time varying mechanical properties of the brain tissue-microelectrode interface due to the hyperelastic, viscoelastic nature of brain tissue b) non-stationarities in the neural signal caused by mechanical and physiological events in the interface and c) the lack of visual feedback of microelectrode position in brain tissue. A closed loop control algorithm is proposed here for autonomous navigation of microelectrodes in brain tissue while optimizing the signal-to-noise ratio of multi-unit neural recordings. The algorithm incorporates a quantitative understanding of constitutive mechanical properties of soft viscoelastic tissue like the brain and is guided by models that predict stresses developed in brain tissue during movement of the microelectrode. An optimal movement strategy is developed that achieves precise positioning of microelectrodes in the brain by minimizing the stresses developed in the surrounding tissue during navigation and maximizing the speed of movement. Results of testing the closed-loop control paradigm in short-term rodent experiments validated that it was possible to achieve a consistently high quality SNR throughout the duration of the experiment. At the systems level, new generation of MEMS actuators for movable microelectrode array are characterized and the MEMS device operation parameters are optimized for improved performance and reliability. Further, recommendations for packaging to minimize the form factor of the implant; design of device mounting and implantation techniques of MEMS microelectrode array to enhance the longevity of the implant are also included in a top-down approach to achieve a reliable brain interface.
ContributorsAnand, Sindhu (Author) / Muthuswamy, Jitendran (Thesis advisor) / Tillery, Stephen H (Committee member) / Buneo, Christopher (Committee member) / Abbas, James (Committee member) / Tsakalis, Konstantinos (Committee member) / Arizona State University (Publisher)
Created2013
Description
Most data cleaning systems aim to go from a given deterministic dirty database to another deterministic but clean database. Such an enterprise pre–supposes that it is in fact possible for the cleaning process to uniquely recover the clean versions of each dirty data tuple. This is not possible in many cases, where the most a cleaning system can do is to generate a (hopefully small) set of clean candidates for each dirty tuple. When the cleaning system is required to output a deterministic database, it is forced to pick one clean candidate (say the "most likely" candidate) per tuple. Such an approach can lead to loss of information. For example, consider a situation where there are three equally likely clean candidates of a dirty tuple. An appealing alternative that avoids such an information loss is to abandon the requirement that the output database be deterministic. In other words, even though the input (dirty) database is deterministic, I allow the reconstructed database to be probabilistic. Although such an approach does avoid the information loss, it also brings forth several challenges. For example, how many alternatives should be kept per tuple in the reconstructed database? Maintaining too many alternatives increases the size of the reconstructed database, and hence the query processing time. Second, while processing queries on the probabilistic database may well increase recall, how would they affect the precision of the query processing? In this thesis, I investigate these questions. My investigation is done in the context of a data cleaning system called BayesWipe that has the capability of producing multiple clean candidates per each dirty tuple, along with the probability that they are the correct cleaned version. I represent these alternatives as tuples in a tuple disjoint probabilistic database, and use the Mystiq system to process queries on it. This probabilistic reconstruction (called BayesWipe–PDB) is compared to a deterministic reconstruction (called BayesWipe–DET)—where the most likely clean candidate for each tuple is chosen, and the rest of the alternatives discarded.
ContributorsRihan, Preet Inder Singh (Author) / Kambhampati, Subbarao (Thesis advisor) / Liu, Huan (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2013
Description
As the size and scope of valuable datasets has exploded across many industries and fields of research in recent years, an increasingly diverse audience has sought out effective tools for their large-scale data analytics needs. Over this period, machine learning researchers have also been very prolific in designing improved algorithms which are capable of finding the hidden structure within these datasets. As consumers of popular Big Data frameworks have sought to apply and benefit from these improved learning algorithms, the problems encountered with the frameworks have motivated a new generation of Big Data tools to address the shortcomings of the previous generation. One important example of this is the improved performance in the newer tools with the large class of machine learning algorithms which are highly iterative in nature. In this thesis project, I set about to implement a low-rank matrix completion algorithm (as an example of a highly iterative algorithm) within a popular Big Data framework, and to evaluate its performance processing the Netflix Prize dataset. I begin by describing several approaches which I attempted, but which did not perform adequately. These include an implementation of the Singular Value Thresholding (SVT) algorithm within the Apache Mahout framework, which runs on top of the Apache Hadoop MapReduce engine. I then describe an approach which uses the Divide-Factor-Combine (DFC) algorithmic framework to parallelize the state-of-the-art low-rank completion algorithm Orthogoal Rank-One Matrix Pursuit (OR1MP) within the Apache Spark engine. I describe the results of a series of tests running this implementation with the Netflix dataset on clusters of various sizes, with various degrees of parallelism. For these experiments, I utilized the Amazon Elastic Compute Cloud (EC2) web service. In the final analysis, I conclude that the Spark DFC + OR1MP implementation does indeed produce competitive results, in both accuracy and performance. In particular, the Spark implementation performs nearly as well as the MATLAB implementation of OR1MP without any parallelism, and improves performance to a significant degree as the parallelism increases. In addition, the experience demonstrates how Spark's flexible programming model makes it straightforward to implement this parallel and iterative machine learning algorithm.
ContributorsKrouse, Brian (Author) / Ye, Jieping (Thesis advisor) / Liu, Huan (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2014