Doctor of Philosophy, The Ohio State University, 2020, Linguistics

This thesis treats a major challenge for current state-of-the-art natural language processing (NLP) pipelines: morphologically rich languages where many inflected forms or weak form--meaning correspondence lead to data sparsity and noise. For example, if the lexeme TEACHER occurs the same number of times in an English text and an Arabic text, those occurrences will be spread over just four forms in English, teacher, teacher's, teachers' and teachers, versus numerous forms in Arabic, leading to more low frequency and out-of-vocabulary forms at test time. Furthermore, while the +s suffix of teachers is highly predictable, there is significant entropy involved in predicting how pluralization will realize in Arabic, which can cause models to be noisy. That said, the particular means of realizing pluralization (among other properties) can be informative in Arabic, as the +wn in mdrswn, 'teachers' not only indicates plurality, but also that the referent is human. To address data sparsity and noise from morphological richness, I propose some practical means of inducing morphological information and/or incorporating morphological information in preprocessing steps or model components, depending on the task at hand. The goals of this intervention are twofold. First, I aim to link variant inflections of the same lexeme to reduce sparsity. Second, I aim to mitigate noise by identifying morphosyntactic properties encoded in complex inflections like mdrswn and leverage them to help models interpret low frequency or out-of-vocabulary forms. To be practical, morphological modeling should be maximally language agnostic, i.e., portable to new languages or domains with minimal human effort, and maximally cheap, i.e., in terms of the amount/cost of required manual supervision. Thus, I explore morphological modeling strategies and morphological resource creation, progressing toward more language agnostic solutions requiring less supervision over the course of this thesis. To (open full item for complete abstract)

Committee: Marie-Catherine de Marneffe (Advisor); Micha Elsner (Committee Member); Nizar Habash (Committee Member); Andrea Sims (Committee Member) Subjects: Computer Science; Linguistics

BA, Oberlin College, 2018, Mathematics

At its core, cryptography relies on problems that are simple to construct but difficult to solve unless certain information (the “key”) is known. Many of these problems come from number theory and group theory. One method of obtaining groups from which to build cryptosystems is to define algebraic curves over finite fields and then derive a group structure from the set of points on those curves. This thesis serves as an exposition of Elliptic Curve Cryptography (ECC), preceded by a discussion of some basic cryptographic concepts and followed by a glance into one generalization of ECC: cryptosystems based on hyperelliptic curves.

Committee: Benjamin Linowitz (Advisor) Subjects: Computer Science; Mathematics

Doctor of Philosophy, The Ohio State University, 2017, Linguistics

Humans are sensitive to the frequency of events, and this sensitivity is reflected in a wide range of behavioral and neural measures. This thesis focuses on the ways in which syntactic co-occurrence frequencies affect human language comprehension. Previous psycholinguistic findings seemed to show that humans are not sensitive to verbal subcategorization frequencies. Instead, this work demonstrates that sensitivity to fine-grained syntactic frequencies provide a confounding explanation for those findings. A left-corner parser is defined that can be used to compute a variety of psycholinguistic complexity metrics in order to better control for such syntactic influences in future studies. One of the strongest and most commonly used psycholinguistic measures output by the parser is surprisal (Hale, 2001; Levy, 2008), which estimates frequency-based comprehension difficulty based on the probability of an observation conditioned on the observations that preceded it. When used to predict reading times, however, this work shows that surprisal is mathematically inconsistent since it conditions on the immediately adjacent lexical material despite the fact that reading proceeds via saccades over non-adjacent material. This mathematical problem with surprisal can be corrected by summing surprisal over each saccade region to enable the measure to account for the probability of each new span of text conditioned on the preceding material that was actually observed. The corrected version of lexical (n-gram) surprisal, cumulative n-gram surprisal, obtains a better fit to reading times than the uncorrected version, though the correction does not work for surprisal over syntactic (probabilistic context-free; PCFG) structure. In addition to the frequency of observed events, this work explores the influence of frequency in how humans predict upcoming events. In particular, uncertainty about upcoming material (entropy) is shown to influence reading times, corroborating previous (open full item for complete abstract)

Committee: William Schuler (Advisor); Micha Elsner (Committee Member); Shari Speer (Committee Member); Shravan Vasishth (Committee Member) Subjects: Computer Science; Linguistics; Psychology

Master of Mathematical Sciences, The Ohio State University, 2017, Mathematics

This project introduces a flexible mathematical model used to represent problems of separating objects in space. Specifically, the notion of disentangling objects is formalized, and the definitions of this model are adapted from several used in the study of topology. Since spatial separation problems have been studied using computational models, methods were considered for translating problems from the mathematical model to that of computation. Upon acknowledging the assumptions made between the models, there is a review of computational complexity results for both determining the ability to disentangle objects and finding the fastest untangling motion.

Committee: Sergei Chmutov PhD (Advisor); Jenny Sheldon PhD (Advisor); Thomas Kerler PhD (Committee Member) Subjects: Mathematics

Master of Science in Engineering, University of Akron, 2014, Electrical Engineering

Wideband beamforming is an essential requirement for several areas of importance, such as radar, cognitive radio and wireless communications. Emerging ultra-wideband (UWB) systems demand rapid electronic steerability, and multiple beams, while maintaining a relatively constant far-field beamwidth. Traditional phased-array based coherent summing approaches generally fail to comply with the requirements of the UWB systems, and implementing such architectures on a hardware level is computationally intensive. Moreover, digital signal processing (DSP) of wideband signals require sophisticated front-ends for pre-processing stages, followed by analog-to-digital converters (ADCs) operating at a very high frequency of operation. In this study, a novel approach to mitigate the ADC requirements is explored, by using both microwave channelizers and sub-sampling ADCs. The microwave channelizer splits the incoming wideband signals into narrower bands, which are then digitized using subsampling ADCs. The resulting signals are simultaneously down-sampled and down-converted. This approach is investigated for a planar manifold-type microwave channelizer, integrated with multi-dimensional filters, to aid 2-D and 3-D wideband directional beamforming systems. The proposed 3-D architecture was primarily investigated for focal plane array (FPA) applications, in order to exploit the high gain of a parabolic dish reflector, while achieving support for multiple electronically scanned beams, by replacing existing horn antennas with an FPA placed at the focal point of the dish. Such architectures, with the aid of the recently proposed frustum shaped filters, find many applications in radar, radio astronomy and other microwave imaging techniques. A novel 2-D bio-inspired spatially band-pass multi-beam filter, from the 1-D electrical equivalent of a mammalian cochlea, is explored. The proposed architectures were evaluated for performance under various scenarios involving wideband directional int (open full item for complete abstract)

Committee: Arjuna Madanayake Dr. (Advisor); Nghi Tran Dr. (Committee Member); Ryan Toonen Dr. (Committee Member) Subjects: Computer Engineering; Electrical Engineering

Doctor of Philosophy (PhD), Ohio University, 2013, Mechanical and Systems Engineering (Engineering and Technology)

This dissertation develops efficient methods for calculating the makespan of a perturbed job shop. All iterative scheduling algorithms require their performance measure, usually the makespan, to be calculated during every iteration. Therefore, this work can enhance the efficiency of many existing scheduling heuristics, e.g. Tabu Search, Genetic Algorithms, Simulated Annealing etc. This increased speed provides two major benefits. The first is the capability of searching a larger solution space, and second is the capability to find a better solution due to the extra time. The following is a list of major highlights of this dissertation. The dissertation extends the hierarchical block diagram model formulation and composition that was originally proposed by Imaev[2]. An algorithm is developed that reduces the complexity of calculating the makespan of the perturbed schedule of job shop with no recirculation from O(MNlogMN) to O(N^2), where M is the number of machines and N the number of parts. An efficient algorithm that calculates kleene star of a lower triangular matrix is presented. This algorithm has complexity of O((n^3)/6) which is 1/16th of the traditional approach. Finally, a novel pictorial methodology, called the SBA (Serial Block Addition), is developed to calculate the makespan of a perturbed job shop. A very efficient single perturbed machine scheduling algorithm, with complexity of O(N^2), is derived using the SBA method. The algorithm was tested on 10,000 randomly generated problems. The solutions provided by scheduling algorithm were 95.27% times, within a 3% deviation of the optimal solutions.

Committee: Robert Judd (Advisor) Subjects: Engineering; Industrial Engineering; Mathematics

Master of Science in Electrical Engineering, Cleveland State University, 2010, Fenn College of Engineering

The increased usage of mobile communication devices has imposed a challenge of achieving efficient communication with minimum power consumption. Moreover, with the advent of software defined radios (SDR), it is highly possible that signal processing functions would be implemented in software in future mobile devices. Hence, the power consumption of these future devices will be directly related to the power consumed by the processor that executes SDR software. This thesis aims at analyzing the computational complexity of different modulation schemes and signal processing communication functions of IEEE WiFi standard. This analysis provides good insight on how the computational load varies at different data rates for different modulation schemes. For this purpose, we have analyzed computational complexity of various modulation schemes and other communication functions using widely known software radio platform i.e. USRP hardware and GNU Radio open source software platform, Matlab and OProfile (open source Linux profiling tool). After performing an extensive analysis, we are able to determine how different modulation schemes and communication functions perform computationally on a given platform. This analysis would help to achieve effective communication along with the efficient use of power in SDR based systems.

Committee: Chansu Yu PhD (Committee Chair); Wenbing Zhao PhD (Committee Member); Fuqin Xiong PhD (Committee Member) Subjects: Computer Engineering; Electrical Engineering