Browsing by Subject "data analytics"

Now showing 1 - 7 of 7
  • Thumbnail Image
    Item
    Data analytics for modeling and visualizing attack behaviors: A case study on SSH brute force attacks
    (IEEE, 2017-11) Yao, Chengchao; Luo, Xiao; Zincir-Heywood, A. Nur; Computer and Information Science, School of Science
    In this research, we explore a data analytics based approach for modeling and visualizing attack behaviors. To this end, we employ Self-Organizing Map and Association Rule Mining algorithms to analyze and interpret the behaviors of SSH brute force attacks and SSH normal traffic as a case study. The experimental results based on four different data sets show that the patterns extracted and interpreted from the SSH brute force attack data sets are similar to each other but significantly different from those extracted from the SSH normal traffic data sets. The analysis of the attack traffic provides insight into behavior modeling for brute force SSH attacks. Furthermore, this sheds light into how data analytics could help in modeling and visualizing attack behaviors in general in terms of data acquisition and feature extraction.
  • Thumbnail Image
    Item
    Data-To-Question Generation Using Deep Learning
    (IEEE, 2023) Koshy, Nicole; Dixit, Anshuman; Jadhav, Siddhi Shrikant; Penmatsa, Arun V.; Samanthapudi, Sagar V.; Kumar, Mothi Gowtham Asok; Anuyah, Sydney Oghenetega; Vemula, Gourav; Herzog, Patricia Snell; Bolchini, Davide
    Many publicly available datasets exist that can provide factual answers to a wide range of questions that benefit the public. Indeed, datasets created by governmental and non- governmental organizations often have a mandate to share data with the public. However, these datasets are often underutilized by knowledge workers due to the cumbersome amount of expertise and embedded implicit information needed for everyday users to access, analyze, and utilize their information. To seek solutions to this problem, this paper discusses the design of an automated process for generating questions that provide insight into a dataset. Given a relational dataset, our prototype system architecture follows a five-step process from data extraction, cleaning, pre-processing, entity recognition using deep learning, and questions formulation. Through examples of our results, we show that the questions generated by our approach are similar and, in some cases, more accurate than the ones generated by an AI engine like ChatGPT, whose question outputs while more fluent, are often not true to the facts represented in the original data. We discuss key limitations of our approach and the work to be done to bring to life a fully generalized pipeline that can take any data set and automatically provide the user with factual questions that the data can answer.
  • Thumbnail Image
    Item
    Does Bad News Spread Faster?
    (IEEE, 2017-01) Fang, Anna; Ben-Miled, Zina; Electrical and Computer Engineering, School of Engineering and Technology
    Bad news travels fast. Although this concept may be intuitively accepted, there has been little evidence to confirm that the propagation of bad news differs from that of good news. In this paper, we examine the effect of user perspective on his or her sharing of a controversial news story. Social media not only offers insight into human behavior but has also developed as a source of news. In this paper, we define the spreading of news by tracking selected tweets in Twitter as they are shared over time to create models of user sharing behavior. Many news events can be viewed as positive or negative. In this paper, we compare and contrast tweets about these news events among general users, while monitoring the tweet frequency for each event over time to ensure that news events are comparable with respect to user interest. In addition, we track the tweets of a controversial event between two different groups of users (i.e., those who view the event as positive and those who view it as negative). As a result, we are able to make assessments based on a single event from two different perspectives.
  • Thumbnail Image
    Item
    Feeding the World with Data: Visions of Data-Driven Farming
    (ACM, 2019-06) Steup, Rosemary; Dombrowski, Lynn; Su, Norman Makoto; Human-Centered Computing, School of Informatics and Computing
    Recent years have seen increased investment in data-driven farming through the use of sensors (hardware), algorithms (software), and networking technologies to guide decision making. By analyzing the discourse of 34 startup company websites, we identify four future visions promoted by data-driven farming startups: the vigilant farmer who controls all aspects of her farm through data; the efficient farmer who has optimized his farm operations to be profitable and sustainable; the enlightened farmer who achieves harmony with nature via data-driven insights; and the empowered farmer who asserts ownership of her farm's data, and uses it to benefit herself and her fellow farmers. We describe each of these visions and how startups propose to achieve them. We then consider some consequences of these visions; in particular, how they might affect power relations between the farmer and other stakeholders in agriculture--farm workers, nonhumans, and the technology providers themselves.
  • Thumbnail Image
    Item
    Introduction
    (Project Muse, 2019) Jones, Kyle M. L.; Library and Information Science, School of Informatics and Computing
  • Thumbnail Image
    Item
    Library Assessment and Data Analytics in the Big Data Era: Practice and Policies
    (Wiley, 2015) Chen, Hsin-liang; Doty, Philip; Mollman, Carol; Niu, Xi; Yu, Jen-chien; Zhang, Tao; Department of Library and Information Science, School of Informatics and Computing
    Emerging technologies have offered libraries and librarians new ways and methods to collect and analyze data in the era of accountability to justify their value and contributions. For example, Gallagher, Bauer and Dollar (2005) analyzed the paper and online journal usage from all possible data sources and discovered that users at the Yale Medical Library preferred the electronic format of articles to the print version. After this discovery, they were able to take necessary steps to adjust their journal subscriptions. Many library professionals advocate such data-driven library management to strengthen and specify library budget proposals.
  • Thumbnail Image
    Item
    Nutritional Assessment of Denture Wearers Using Matched Electronic Dental-Health Record Data
    (Wiley, 2022-08) Felix Gomez, Grace Gomez; Cho, Sopanis D.; Varghese, Roshan; Rajendran, Divya; Eckert, George J.; Bhamidipalli, Sruthi Surya; Gonzalez, Theresa; Khan, Babar Ali; Thyvalikakath, Thankam Paul; Cariology, Operative Dentistry and Dental Public Health, School of Dentistry
    Purpose To assess the nutritional profile of denture wearers through a retrospective cohort study using nutritional biomarkers from matched electronic dental and health record (EDR-EHR) data. Materials and methods The case group (denture wearers) included matched EDR-EHR data of patients who received removable partial, complete, and implant-supported prosthodontic treatments between January 1, 2010 and December 31, 2018, study time. The control (nondenture wearers) group did not have recorded denture treatments and included patient records within 1 year of the denture index date (first date of case patients’ receiving complete or partial denture) of the matching cases. The qualified patients’ EDR were matched with their EHR based on the availability of laboratory reports within 2 years of receiving the dentures (index date). Nutritional biomarkers were selected from laboratory reports for complete blood count, comprehensive and basic metabolic profile, lipid, and thyroid panels. Summary statistics were performed, and general linear mixed effect models were used to evaluate the rate of change over time (slope) of nutritional biomarkers before and after the index date. Likelihood ratio tests were performed to determine the differences between dentures and controls. Results The final cohort included 10,481 matched EDR-EHR data with 3,519 denture wearers and 6,962 controls that contained laboratory results within the study time. The denture wearers’ mean age was 57 ±10 years and the control group was 56 ±10 years with 55% females in both groups. Pre-post analysis among denture wearers revealed decreased serum albumin (p = 0.002), calcium (p = 0.039), creatinine (p < 0.001) during the post-index time. Hemoglobin (Hb) was higher pre-index, and was decreasing during the time period but did not change post-index (p < 0.001). Among denture wearers, completely edentulous patients had a significant decrease in serum albumin, creatinine, blood urea nitrogen (BUN), but increased estimated glomerular filtration rate (eGFR). In partially edentulous patients, total cholesterol decreased (p = 0.018) and TSH (p = 0.004), BUN (p < 0.001) increased post-index. Patients edentulous in either upper or lower arch had decreased BUN and eGFR during post-index. Compared to controls, denture wearers showed decreased serum albumin and protein (p = 0.008), serum calcium (p = 0.001), and controls showed increased Hb (p = 0.035) during post-index. Conclusions The study results indicate nutritional biomarker variations among denture wearers suggesting a risk for undernutrition and the potential of using selected nutritional biomarkers to monitor nutritional profile.