Patrick Donnelly

In recent years, there has been a renewed interest in languages and systems for large scale distributed computing. Unfortunately, most systems available to the end user use a custom description language tightly coupled to a specific runtime implementation, making it difficult to transfer applications between systems. To address this problem we introduce Makeflow, a simple system for expressing and running a data-intensive workflow across multiple execution engines without requiring changes to the application or workflow description. Makeflow allows any user familiar with basic Unix Make syntax to generate a workflow and run it on one of many supported execution systems. Furthermore, in order to assess the performance characteristics of the various execution engines available to users and assist them in selecting one for use we introduce Workbench, a suite of benchmarks designed for analyzing common workflow patterns. We evaluate Workbench on two physical architectures -- the first a storage cluster with local disks and a slower network and the second a high performance computing cluster with a central parallel filesystem and fast network -- using a variety of execution engines. We conclude by demonstrating three applications that use Makeflow to execute data intensive applications consisting of thousands of jobs.

We investigate automatic analysis of teachers' instructional strategies from audio recordings collected in live classrooms. We collected a data set of teacher audio and human-coded instructional activities (e.g., lecture, question and answer, group work) in 76 middle school literature, language arts, and civics classes from eleven teachers across six schools. We automatically segment teacher audio to analyze speech vs. rest patterns, generate automatic transcripts of the teachers' speech to extract natural language features, and compute low-level acoustic features. We train supervised machine learning models to identify occurrences of five key instructional segments (Question & Answer, Procedures and Directions, Supervised Seatwork, Small Group Work, and Lecture) that collectively comprise 76% of the data. Models are validated independently of teacher in order to increase generalizability to new teachers from the same sample. We were able to identify the five instructional segments above chance levels with F1 scores ranging from 0.64 to 0.78. We discuss key findings in the context of teacher modeling for formative assessment and professional development.

We investigate automatic detection of teacher questions from audio recordings collected in live classrooms with the goal of providing automated feedback to teachers. Using a dataset of audio recordings from 11 teachers across 37 class sessions, we automatically segment the audio into individual teacher utterances and code each as containing a question or not. We train supervised machine learning models to detect the human-coded questions using high-level linguistic features extracted from automatic speech recognition (ASR) transcripts, acoustic and prosodic features from the audio recordings, as well as context features, such as timing and turn-taking dynamics. Models are trained and validated independently of the teacher to ensure generalization to new teachers. We are able to distinguish questions and non-questions with a weighted F1 score of 0.69. A comparison of the three feature sets indicates that a model using linguistic features outperforms those using acoustic-prosodic and context features for question detection, but the combination of features yields a 5% improvement in overall accuracy compared to linguistic features alone. We discuss applications for pedagogical research, teacher formative assessment, and teacher professional development.

The main objective of the VIVA project (Visual Identity Verification Auditor) is to investigate and demonstrate a professional broadcast surveillance system. Broadcast material is pre-encoded with an invisible and unique watermark identifier. By automatically monitoring television broadcasts and registering which assets have been pre-encoded, a mechanism for IPR protection is provided. The applications of such a system include copyright protection and proof of ownership, verification of commercial transmissions, assessment of sponsorship effectiveness, protection against illegal transmission, statistical data collection and analysis of broadcast content. The watermarking technology is optimised for the high picture quality needed in a broadcast environment. At the same time, the watermark survives signal processing operations which routinely occur in broadcasting systems such as digital to analogue conversion, editing and compression. The monitoring system detects a 36-bit payload every second, which guarantees an operation time of more than 13 years, without recourse to repeat identifiers. The detection algorithm has reasonably low complexity, enabling real time watermark detection for many broadcast channels simultaneously. We report on the first results of a field trial, using a satellite link between Sweden and Belgium, proving the feasibility of the system.

In music, multiple pitches often occur simultaneously, an essential feature of harmony. In the present study, the authors assessed the ability of cochlear implant (CI) users to perceive polyphonic pitch. Acoustically presented stimuli consisted of one, two, or three superposed tones with different fundamental frequencies (f(0)). The normal hearing control group obtained significantly higher mean scores than the CI group. CI users performed near chance levels in recognizing two- and three-pitch stimuli, and demonstrated perceptual fusion of multiple pitches as single-pitch units. These results suggest that limitations in polyphonic pitch perception may significantly impair music perception in CI users.