黄色直播

听

Visual Analytics Research in Computer Science

A major project is underway in the Faculty of Computer Science in Visual Analytics, funded by the Boeing Company. The project started in October 2010. The project has two parts, Visual Text Analytics (over 5 years) and Mobile Graphics (over 3 years).

Visual text analytics

Visual text analytics aims to advance the state of the art in sense-making of large document corpora. The search paradigm for finding information of interest in text collections is well established, and exemplified by the success of Web search engines. The next frontier is the support of sense-making out of medium to large scale text corpora by the domain expert and analyst, who is trying to untap the tacit knowledge hidden in the text. Visual text analytics brings together text mining, text visualization and human computer interaction to create computer tools to support this activity. Key challenges include: support for real-time visualization and interaction, novel text visualization techniques (possibly domain dependent) and interaction techniques that permit the domain analyst to fine tune the text mining, without becoming a text mining expert themself.

The Boeing and Aeroinfo听Systems Projects

Boeing and one of its Canadian affiliates, Aeroinfo Systems, have a major interest in the application of visual text analytics to the mining of aviation safety related corpora aiming to improve all aspects of safety in aircraft manufacturing, operation and maintenance. This is the main driver of the proposed project. Additional domains that will be addressed, to broaden the scope of the research, and the generality of the proposed methodologies, include: help desk application (trouble ticket mining); Community Question-Answering forum mining; research literature and patent mining; financial report and time series mining.听 The project will train a new generation of students, who will bridge the areas of text mining, text visualization and human computer interaction, to support current and future needs of this growing Canadian industry sector. The project involves several faculty members, including Vlado Keselj, Mike Shepherd, Stephen Brooks, Kirstie Hawkey and Raza Abidi, and two postdoctoral researchers, Axel Soto and Aminul Islam. A strong team of graduate students, both at the PhD level (Magda Jankowska, Armin Sajadi, Raheleh Makki Niri) and at the Master鈥檚 level (Shali Liu, Tomasz Niewiariowski) started their degrees in September 2011, and are already producing some interesting work through projects in their directed study on Visual Text Analytics this fall. Undegraduate student Ryan Kiros worked on the project under an NSERC USRA in the summer of 2011. The project has acquired 6 desktop computers running Linux, tied together into a Hadoop cluster by PhD student Marek Lipczak. The project has access to the Compute Canada AceNet high-performance computing facility, and a powerful computing cluster in the Faculty of Computer Science.

Mobile Graphics

The Mobile Graphics part of the project aims to support airplane mechanics, who work in a very complex physical and digital world whether they are in the factory assembling planes or conducting maintenance work at airports. Their digital workspace is accessed through multiple devices, including printouts. At the start of each shift on the factory floor, mechanics view the instructions and drawings of the aircraft that they will be working during that shift. There is a stationary workstation on the shop floor where they can print paper engineering drawings and specs in order to take both the appropriate tools and proper protocols to the aircraft. Once at the aircraft, they have two options to access the information on site: using a stationary display terminal and printer at the assembly station outside the aircraft or sign out a laptop computer for use on the aircraft. The stationary workstation, of course, leads to unwieldy paper printouts and/or multiple interruptions to work flow. The ruggedized laptops are too cumbersome in many situations, such as on a ladder or in a confined space. The introduction of newer smaller mobile devices, while appealing, has constraints on screen size, interaction techniques, and complexity of mental model. The project investigates different aspects of the overall problem of supporting aircraft mechanics during information retrieval tasks in complex information spaces that include text, 2D and 3D images. As the mechanics also must orient themselves in physical space and integrate the digital realm with real-world parts, augmented reality (AR), including 3D vision and an RFID system, may aid them in their tasks. The project involves faculty members Kirstie Hawkey, Derek Reilly, Qigang Gao, Carolyn Watters, Stephen Brooks, Dirk Arnold and Srini Sampalli, and postdoctoral researcher Bonnie MacKay. Several graduate students working on the project started their degrees in September 2011, including PhD student Raghav Sampangi, and Master鈥檚 students Xiaoyu Yu, Ankur Sharma and Matt MacDonald. Two undergraduate students worked on setting up the lab facilities for the Mobile Graphics project in the summer of 2011. The lab is equipped with large screen displays and an assortment of mobile devices.

In the fall of 2011, the Killam Lecture Series focused on the Visual Analytics theme, and hosted two internationally prominent researchers, Daniel Keim from the University of Konstanz in Germany and John Stasko from the Georgia Institute of Technology.

If all goes as expected with matching the Boeing funds with the generous support of the Federal Government through jointly funded NSERC collaborative research and development projects, MITACS internships and NSERC Industrial Postgraduate scholarships, we hope to double the number of students in the coming year, and establish 黄色直播 Computer Science as a hub of Visual Analytics in Eastern Canada.