Theses and Dissertations from UMD
Permanent URI for this communityhttp://hdl.handle.net/1903/2
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a give thesis/dissertation in DRUM
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
3 results
Search Results
Item APPROACHING GENDER PARITY: WOMEN IN COMPUTER SCIENCE AT AFGHANISTAN'S KABUL UNIVERSITY(2010) Plane, Jandelyn Dawn; Selden, Steven; Education Policy, and Leadership; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This study explores the representation of women in computer science at the tertiary level through data collected about undergraduate computer science education at Kabul University in Afghanistan. Previous studies have theorized reasons for underrepresentation of women in computer science, and while many of these reasons are indeed present in Afghanistan, they appear to hinder advancement to degree to a lesser extent. Women comprise at least 36% of each graduating class from KU's Computer Science Department; however, in 2007 women were 25% of the university population. In the US, women comprise over 50% of university populations while only graduating on average 25% women in undergraduate computer science programs. Representation of women in computer science in the US is 50% below the university rate, but at KU, it is 50% above the university rate. This mixed methods study of KU was conducted in the following three stages: setting up focus groups with women computer science students, distributing surveys to all students in the CS department, and conducting a series of 22 individual interviews with fourth year CS students. The analysis of the data collected and its comparison to literature on university/department retention in Science, Technology, Engineering and Mathematics gender representation and on women's education in underdeveloped Islamic countries illuminates KU's uncharacteristic representation of women in its Computer Science Department. The retention of women in STEM through the education pipeline has several characteristics in Afghanistan that differ from countries often studied in available literature. Few Afghan students have computers in their home and few have training beyond secretarial applications before considering studying CS at university. University students in Afghanistan are selected based on placement exams and are then assigned to an area of study, and financially supported throughout their academic career, resulting in a low attrition rate from the program. Gender and STEM literature identifies parental encouragement, stereotypes and employment perceptions as influential characteristics. Afghan women in computer science received significant parental encouragement even from parents with no computer background. They do not seem to be influenced by any negative "geek" stereotypes, but they do perceive limitations when considering employment after graduation.Item Battle of the Brains: Election-Night Forecasting at the Dawn of the Computer Age(2010) Chinoy, Ira; Beasley, Maurine; Journalism; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This dissertation examines journalists' early encounters with computers as tools for news reporting, focusing on election-night forecasting in 1952. Although election night 1952 is frequently mentioned in histories of computing and journalism as a quirky but seminal episode, it has received little scholarly attention. This dissertation asks how and why election night and the nascent field of television news became points of entry for computers in news reporting. The dissertation argues that although computers were employed as pathbreaking "electronic brains" on election night 1952, they were used in ways consistent with a long tradition of election-night reporting. As central events in American culture, election nights had long served to showcase both news reporting and new technology, whether with 19th-century devices for displaying returns to waiting crowds or with 20th-century experiments in delivering news by radio. In 1952, key players - television news broadcasters, computer manufacturers, and critics - showed varied reactions to employing computers for election coverage. But this computer use in 1952 did not represent wholesale change. While live use of the new technology was a risk taken by broadcasters and computer makers in a quest for attention, the underlying methodology of forecasting from early returns did not represent a sharp break with pre-computer approaches. And while computers were touted in advance as key features of election-night broadcasts, the "electronic brains" did not replace "human brains" as primary sources of analysis on election night in 1952. This case study chronicles the circumstances under which a new technology was employed by a relatively new form of the news media. On election night 1952, the computer was deployed not so much to revolutionize news reporting as to capture public attention. It functioned in line with existing values and practices of election-night journalism. In this important instance, therefore, the new technology's technical features were less a driving force for adoption than its usefulness as a wonder and as a symbol to enhance the prestige of its adopters. This suggests that a new technology's capacity to provide both technical and symbolic social utility can be key to its chances for adoption by the news media.Item Using Locality and Interleaving Information to Improve Shared Cache Performance(2009) Liu, Wanli; Yeung, Donald; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The cache interference is found to play a critical role in optimizing cache allocation among concurrent threads for shared cache. Conventional LRU policy usually works well for low interference workloads, while high cache interference among threads demands explicit allocation regulation, such as cache partitioning. Cache interference is shown to be tied to inter-thread memory reference interleaving granularity: high interference is caused by ne-grain interleaving while low interference is caused coarse-grain interleaving. Proling of real multi-program workloads shows that cache set mapping and temporal phase result in the variation of interleaving granularity. When memory references from dierent threads map to disjoint cache sets, or they occur in distinct time windows, they tend to cause little interference due to coarse-grain interleaving. The interleaving granularity measured by runlength in workloads is found to correlate with the preference of cache management policy: ne-grain interleaving workloads perform better with cache partitioning, and coarse-grain interleaving workloads perform better with LRU. Most existing shared cache management techniques are based on working set locality analysis. This dissertation studies the shared cache performance by taking both locality and interleaving information into consideration. Oracle algorithm which provides theoretical best performance is investigated to provide insight into how to design a better practical policy. Proling and analysis of Oracle algorithm lead to the proposal of probabilistic replacement (PR), a novel cache allocation policy. With aggressor threads information learned on-line, PR evicts the bad locality blocks of aggressor threads probabilistically while preserving good locality blocks of non-aggressor threads. PR is shown to be able to adapt to the different interleaving granularities in different sets over time. Its flexibility in tuning eviction probability also improves fairness among thread performance. Evaluation indicates that PR outperforms LRU, UCP, and ideal cache partitioning at moderate hardware cost. For single program cache management, this dissertation also proposes a novel technique: reuse distance last touch predictor (RD-LTP). RD-LTP is able to capture reuse distance information, which represents the intrinsic memory reference pattern. Based on this improved LT predictor, an MRU LT eviction policy is developed to select the right victim at the presence of incorrect LT prediction. In addition to LT predictor, another predictor: reuse distance predictors (RDPs) is proposed, which is able to predict actual reuse distance values. Compared to various existing cache management techniques, these two novel predictors deliver higher cache performance with higher prediction coverage and accuracy at moderate hardware cost.