Browsing by Type "conference paper"
Now showing 1 - 6 of 6
Results Per Page
Sort Options
Item Event Nugget Detection using Thresholding and Classification Techniques(National Institute of Standards and Technology, 2016-11-14) Satyapanich, Taneeya; Finin, TimThis paper describes the Event Nugget Detection system that we submitted to the TAC KBP 2016 Event Track. We sent out two runs; UMBC1 and UMBC2. UMBC1 is a sentence-level classification system based on Convolution Neural Network and applied the probability to select a word as an event nugget. UMBC2 is the classification model trained from our features using Weka and filtered out low confidence prediction output using threshold. Our performance was low; we got F1 measure of 34.14 for UMBC1 and 35.24 for UMBC2.Item Hash-Grams: Faster N-Gram Features for Classification and Malware Detection(Association for Computing Machinery (ACM), 2018) Raff, Edward; Nicholas, Charles; UMBC Faculty CollectionN-grams have long been used as features for classification problems, and their distribution often allows selection of the top-k occurring n-grams as a reliable first-pass to feature selection. However, this top-k selection can be a performance bottleneck, especially when dealing with massive item sets and corpora. In this work we introduce Hash-Grams, an approach to perform top-k feature mining for classification problems. We show that the Hash-Gram approach can be up to three orders of magnitude faster than exact top-k selection algorithms. Using a malware corpus of over 2 TB in size, we show how Hash-Grams retain comparable classification accuracy, while dramatically reducing computational requirements.Item Performance of Eyes-Free Mobile Authentication(The Internet Society, 2018-02-18) Wolf, Flynn; Aviv, Adam J.; Kuber, RaviMobile device users avoiding observational attacks and coping with situational impairments may employ techniques for eyes-free mobile unlock authentication, where a user enters his/her passcode without looking at the device. This study supplies an initial description of user accuracy in performing this authentication behavior with PIN and pattern passcodes, with varying lengths and visual characteristics. Additionally, we inquire if tactile-only feedback can provide assistive spatialization to support users when interacting with the mobile interface, finding that orientation cues prior to unlocking do not help. A within-group, randomized study was conducted with 26 participants. 1,021 passcode entry gestures were performed under eyes-free conditions. Edit distance measurements were then calculated. Gesture traces and subjective feedback were recorded for subsequent analysis. We found that users who focused on orienting themselves to position the first digit of the passcode using the tactile feedback performed better in the task. These results could be applied to better define eyes-free behavior in further research, and to design better and more secure methods for eyes-free authentication.Item Semantic knowledge and privacy in the physical web(CEUR, 2016-10-18) Das, Prajit Kumar; Kashyap, Abhay L.; Singh, Gurpreet; Matuszek, Cynthia; Finin, Tim; Joshi, AnupamIn the past few years, the Internet of Things has started to become a reality; however, its growth has been hampered by privacy and security concerns. One promising approach is to use Semantic Web technologies to mitigate privacy concerns in an informed, flexible way. We present CARLTON, a framework for managing data privacy for entities in a Physical Web deployment using Semantic Web technologies. CARLTON uses context-sensitive privacy policies to protect privacy of organizational and personnel data. We provide use case scenarios where natural language queries for data are handled by the system, and show how privacy policies may be used to manage data privacy in such scenarios, based on an ontology of concepts that can be used as rule antecedents in customizable privacy policies.Item Statistical Unigram Analysis for Source Code RepositoryXu, Weifeng; Xu, Dianxiang; Ariss, Omar El; Liu, Yunkai; Alatawi, Abdularaham; School of Criminal Justice; Computer ScienceUnigram is a fundamental element of n-gram in natural language processing. However, unigrams collected from a natural language corpus are unsuitable for solving problems in the domain of computer programming languages. In this paper, we analyze the properties of unigrams collected from an ultra-large source code repository. Specifically, we have collected 1.01 billion unigrams from 0.7 million open source projects hosted at GitHub.com. By analyzing these unigrams, we have discovered statistical patterns regarding (1) how developers name variables, methods, and classes, and (2) how developers choose abbreviations. Our study describes a probabilistic model for solving a well-known problem in source code analysis: how to expand a given abbreviation to its original indented word. It shows that the unigrams collected from source code repositories are essential resources to solving the domain specific problems.Item Surface Modification at Nanoscale; Nanoparticle-Nanowire Transition(NASA, 2018-04-15) Singh, N. B.; Su, Ching Hua; Coriell, S. R.; Mandal, K. D.; Arnold, Brad; Choa, Fow-Sen; Cullum, BrianBinary, ternary and quaternary oxides and selenides have been developed and used in multiple applications including high power lasers, detectors, dielectric energy storage and variety of optical devices. These materials have been grown by Bridgman, physical vapor transport (PVT), chemical vapor transport (CVT) methods and flux methods in the form of bulk thin film, nanocrystals and nanowires. With increasing thrust of bio applications, nanoparticles it is essential to understand nucleation and nanomorphological transition during drug delivery, growth of nanoengineered bio composites in body, grain growth and final morphology. Addition of fluorides and selenides have increased significantly in synthetic tissue constituents because of some advantages in adhesion and stability. We have performed experiments on multinary oxides Sr-Ba-O-F, Se-Tl-As and Se-Pb-Sn-Se using several growth methods to demonstrate nanoparticle and nanowire transition. This study has great potential to increase surface area and also provides understanding to the mechanism of nanowire growth.