Foaad Khosmood

User identification through command history analysis

As any veteran of the editor wars can attest, Unix users can be fiercely and irrationally attached to the commands they use and the manner in which they use them. In this work, we investigate the problem of identifying users out of a large set of candidates (25-97) through their command-line histories. Using standard algorithms and feature sets inspired by natural language authorship attribution literature, we demonstrate conclusively that individual users can be identified with a high degree of accuracy through their command-line behavior. Further, we report on the best performing feature combinations, from the many thousands that are possible, both in terms of accuracy and generality. We validate our work by experimenting on three user corpora comprising data gathered over three decades at three distinct locations. These are the Greenberg user profile corpus (168 users), Schonlau masquerading corpus (50 users) and Cal Poly command history corpus (97 users). The first two are well known corpora published in 1991 and 2001 respectively. The last is developed by the authors in a year-long study in 2014 and represents the most recent corpus of its kind. For a 50 user configuration, we find feature sets that can successfully identify users with over 90% accuracy on the Cal Poly, Greenberg and one variant of the Schonlau corpus, and over 87% on the other Schonlau variant.

Gaining efficiency in human assisted transcription and speech annotation in legislative proceedings

We present a study using the Digital Democracy transcription tool. Human transcribers work to up-level and annotate California state legislative proceedings using the tool. Four phases of UI and functionality improvements are introduced and for each phase, the resulting change in efficiency is measured and presented.

Learning alignments from legislative discourse

In this work, we seek to quantify the extent to which a legislators spoken language indicates their degree of alignment toward an organization that has a taken a documented position on some legislation. To perform this study, we use a corpus of bill discussion transcripts provided by Digital Democracy1. We then apply proven learning methods in the field of natural language processing to predict alignment scores between each member of the California state legislature and a select set of state-recognized organizations. Our methods surpass established baselines, achieving up to 78% accuracy when predicting these same scores using discourse features.

3D visualization of legislative relationships

Government relationships can be complex and difficult to understand. The relationships between members of a legislature, bills, votes and lobbyists who promote various causes are important to understand in representative democracies, but difficult to retrieve using current methods. In this paper, we propose a 3D visualization system to explore such legislative relationships for users. We use real data from California state legislature obtained from the Digital Democracy project. We also conduct a 20 person user study to gauge the differences between traditional ways of looking up information versus our graph based methods. While not being as comprehensive, most users found our interactive visualizations more intuitive than regular web-based information retrieval.

predicting the vote using legislative speech

As most dedicated observers of voting bodies like the U.S. Supreme Court can attest, it is possible to guess vote outcomes based on statements made during deliberations or questioning by the voting members. We show this is also possible to do automatically using machine learning, potentially providing a powerful tool to ordinary citizens. Our working hypothesis is that verbal utterances made during the legislative process by elected representatives can indicate their intent on a future vote, and therefore can be used to automatically predict said vote to a significant degree. In this paper, we examine thousands of hours of legislative deliberations from the California state legislatures 2015-2016 session to form models of voting behavior for each legislator and use them to train classifiers and predict the votes that occur subsequent to the discussions. We can achieve average legislator vote prediction accuracies as high as 83%. For bill vote prediction, our model can achieve 76% accuracy with an F1 score of 0.83 using a balanced dataset.

SLO Hacks: Embracing the Passionate Novice

Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the Owner/Author. ICGJ 2018, March 18, 2018, San Francisco, CA, USA 2018 Copyright is held by the owner/author(s). ACM ISBN 978-1-4503-6484-3/18/03. https://doi.org/10.1145/3196697.3196706 Introduction We present this event report and discussion on the SLO Hacks “local hack day”, a 12 hour free-format hackathon held on December 2, 2017 at the Campus of California Polytechnic State University (Cal Poly). A first of its kind for our community, the event presents unique challenges and opportunities. SLO Hacks, an entirely student run volunteer organization, has mobilized to overcome institutional, logistical and educational challenges in order to enable hackathon events at Cal Poly. An analysis of attendees and projects reveals overwhelming support from younger students: teens barely out of high school or recent transfers into Cal Poly. While many are enrolled in technical programs, very few have software or hardware development experience. However, the rapid growth in availability of powerful software tools, such as data analysis and visualization tools, machine learning libraries, combined with an abundance of examples and tutorials on the Internet allow the creation of meaningful projects even by novices, further feeding the passion and excitement around hackathons. We believe the “passionate novice” is a major factor in most hackathon events, and hope to discuss ways to better embrace this class of participant.

Immersive Learning Research Network

This paper describes a multi-user learning environment in a virtual world setting. An exploratory and collaborative learning using an educational scavenger hunt metaphor form the basis for student learning and engagement in virtual world. The learning experience is based on the following elements: exploration – students explore the learning content on their own, and build a knowledge base. Cooperation and Collaboration – students cooperate and collaborate to uncover information, share findings, and gain knowledge and skills. Discussion and Reflection – students discuss and solve problems together, and exercise reflective learning. Based on this idea, three main contributions are provided in this paper: Firstly, a pedagogical model which combines immersive, online, and virtual collaboration with an exploratory teaching approach. Secondly, the learning tasks and interactions are incorporated by a flexible to use set of tools in the virtual world Open Wonderland. Finally, an experimentation study evaluating the virtual world in the learning domain Egyptology.

Inducing Emotional Response in Interactive Media: A Pilot Study

Video games, entertainment, education, and training media have been developed for many years, and eliciting emotional experiences is an integral part of that process. Production and editing of the media in order to produce the desired emotional experiences can be expensive and cumbersome to media designers. This paper presents a pilot study intended to show that such experiences can be induced with after-the-fact audio-visual effects. As subjects of the user study, players are given the same virtual environment with two emotional states: fear, and peace. Over 70 % of players report feeling the proper emotional response in both environmental states, revealing that it is indeed possible to induce emotional response with after-the-fact audio-visual effects, hinting at future possibilities for drag-and-drop emotional experience filters.

Generation of Infotips from Interface Labels

A method is presented for generating informative and natural-sounding infotips for graphical elements of a user interface. A domain-specific corpus is prepared using natural language processing techniques, and a term-frequency/inverse-document-frequency transform is used for vectorization of features. A k-means algorithm is then used to cluster the corpus by semantic similarity and retrieve the most similar infotips for any inputted interface label. We demonstrate the feasibility of this method and conclude by proposing several approaches to improve the selection of infotips by incorporating natural language processing and machine learning techniques.