Bookmark and Share

Speaking: Research

I'm available to speak on a wide range of professional and research topics drawn from my Research areas and have presentations available in a variety of formats from 30 minutes to an entire day and for groups from a small executive roundtable to a theater with several hundred. My Vita contains the most authoritative list of my various conference presentations. I present regularly to industry and governmental groups and can tailor presentations for your specific needs, including non-disclosure. Please contact me for more details.

Several of my popular talks are listed on this page.

How Your Company Can Profit from Social Media: A Short Course

One of my most popular talks, this two-hour shortcourse offers a wide-ranging overview of how the modern company can make the most from social media. Topics covered include internal collaboration, connecting with customers, recruiting, crowdsourcing of ideas (both internally and from customers), use of social media in regulated industries, brand and industry mining, customer interaction, corporate intelligence, crisis communication in the social sphere, and advanced modeling and analysis of social media, predicting the path and impact of coverage of your company in the social environment.

Social media is proving to be a powerful and disruptive force for today's company regardless of industry: more than 50 million Tweets are posted each day, and nearly a quarter of all US internet accesses are to social media sites. Social media can be dangerous if underestimated: a single YouTube video from a disgruntled customer cut Delta Airline's stock by 10% last year, and "telephone game" Twitter posts can escalate a minor incident into a national event. At the same time, social media offers rich potential if used right: Dell's IdeaStorm site has generated more than 10,000 ideas from customers and 400 were so good that the company has already implemented them.

Split into modules, this shortcourse focuses on each topic with time for interaction between each, and draws from a wide range of examples and applications to showcase best practices on social media use today, as well as offer a glimpse into the future of how companies will be able to leverage these new platforms.

KEYWORDS: social media, collaboration, customers, recruiting, crowdsourcing, regulated industries, brand mining, industry mining, customer interaction, crisis communication, competitive intelligence

Analyzing the Global Discourse on Climate Change in Mainstream and Social Media Spheres

The Carbon Capture Report (http://www.carboncapturereport.org/) is a service of the University of Illinois that monitors a daily global sample of English-language news coverage of 15 topic areas including alternative energy, carbon credits, carbon capture and climate change. It performs a wide range of analysis on the content, including clustering multiple copies of the same story in different outlets together, determining the geographic location of the outlet and placing it on Google Maps / Google Earth (to allow one to see how coverage in France is different from India, for example). The site monitors Twitter (90% of climate change tweets contain a link to a web site), YouTube, social bookmarking sites, and the blogosphere, "crowdsourcing" the latest trends from the social sphere and producing an integrated view of the global discourse. We also compile a fully automated bibliographic database that extracts all of the names from each day's news and compiles biographies over time of their coverage patterns.

KEYWORDS: discourse analysis, news analytics, data mining, business intelligence

Cataloging the World: Using Open Source Intelligence and News Reports to Catalog the World's Events 1946-Present

The signature initiative of the Cline Center for Democracy is the Social Political Economic Event Database (SPEED), an effort to compile a global event database covering 172 countries from 1946-present. The resulting database will consist of all major human events in those categories occuring anywhere on earth in the past 60 years, offering an unprecedented view into the underpinnings of societal evolution. More than 350 variables are captured for each event, including date and location to city or landmark resolution. A combination of machine and human processes are used to automate much of the heavy lifting of event coding, while still allowing for the sophistication of interpretation afforded by human analysts. The combined event dataset takes the form a vast network of human activity, with each event being a node and all 350 variables acting as links connecting those events. Predictive analytics and "scenario modeling" can ultimately use this event network to forecast sets of potential outcomes for emerging conflict across the globe.

KEYWORDS: data mining, event databases, news analysis, open source intelligence, intelligence

Computerized (NLP) Approaches to Event Extraction From Textual Reports

Ultimately every crisis mapping system relies on a set of codified data with locations and attribute information about a set of events. Much of the realtime and archival data for crisis mapping comes not from well-structured forms entry with controlled vocabularies, but from adhoc reports either through the news media or self-submitted reports from the field. Processing these textual reports to extract event records with date, location, and attribute information has required a number of advances in the state-of-the-art in computer text processing, from disambiguating geocoders that can "reason" about spatial information based on context, to significant advances in semantic processing and natural language processing. This talk surveys many of the aspects of modern automated event mining systems.

KEYWORDS: web mining, data mining, event databases, news analysis, open source intelligence, intelligence

Open Sources and Foreign Media Analysis

Open Source Intelligence (OSI) refers to the use of "open sources" such as newspapers, magazines, television, radio, and other public news outlets to collect intelligence and perform analysis on a particular subject. Foreign Media Analysis is a component of open source analysis involved in translation or vernacular foreign media content and deriving analytical intelligence products ranging from factual elements of the content to the discourse elements of tone, structure, and language use.

KEYWORDS: open source intelligence, foreign media analysis, content analysis, translation

Collaborative Research In A "Big Data" World: Harnessing Advanced Collaborative Technology For Large-Scale Research

The academic and business worlds continue on their path towards "big data" research, in which large geographically distributed teams, massive datasets, and huge computational resources are brought together with advanced data management, spatial and temporally aware data systems, network analysis systems, and advanced visualization tools to meet the demands of next-generation research. This talk provides an overview of the unique environment the Cline Center for Democracy has built to support its massive-scale data projects, including "awareness-enabled" communication tools that can proactively and intelligently analyze communication flows, data repositories designed for massively parallel collaboration, protocol systems for human-computer collaborative document coding of hundreds of millions of documents, and state-of-the-art visualization and interpretation tools.

KEYWORDS: data mining, collaboration, user interfaces, massive data, network analysis, visualization, awareness

Digital Media Management and Large-Scale Digitization and Digital Access: the Phantasm and UIHistories Projects

We live in a media-rich world where every cellphone is a camera and visitors to Facebook alone upload more than two billion photographs each month. How can we manage that kind of media overload? How can the "mainstreaming" of digital media be used to document the day-to-day evolution of an organization? How can rights-management technology automate the access-control aspects of digital media? Digitization and digital history have seen explosive growth in recent years with the advent of bulk-digitization workflows like Google Books and the Open Content Alliance. What can we learn from these processes to help guide our own digitization efforts? In particular, what kinds of interface modalities seem to work the best for allowing users to interact with digital content, and how can digital document repositories be made more interactive? This talk will focus on experiences learned from the Phantasm project (a collection of more than 80,000 photographs of the University of Illinois) and the UIHistories project (more than 300 historical narratives combined with more than 70,000 pages of digitized material), in which five years of research have led to countless lessons learned that I will be sharing in this talk, along with a discussion of the current technology platforms that were developed to support these projects.

KEYWORDS: digital media, media delivery platforms, digital access, digitization

Cataloging a Campus - Information management lessons from the Phantasm project, a collection of more than 80,000 images of the Universiy of Illinois

The Phantasm project http://uihistoriesproject.chass.illinois.edu/photoarchive/ is a collection of more than 80,000 images of the University of Illinois at Urbana-Champaign campus, covering every major structure and space of this institution. For more than five years, the Phantasm project has documented campus through the four seasons. Groundbreakings, open houses, student events, and even everyday life have been captured in this archive for future generations. The project was created to set the groundwork for the creation of a single campus photographic repository that would address the immediate needs of campus for a comprehensive collection of campus imagery, while at the same time, providing a unique "photographic time capsule" for the future.

Beyond its usefulness as a source of images, the project also presents a number of interesting problems in indexing such a large collection of images and making it accessible. How exactly does one create a taxonomy to describe the buildings of a University campus and their contents? What are the issues involved with having a single index that must support identification ranging from light fixtures to buildings? How can a single index accommodate a widely varied set of user communities, ranging from campus historians to marketing and communications staff? The very high accretion rate of the project (in excess of 1,000-5,000 images a day at peak) necessitates a very flexible and optimized workflow, and the NCSA Phantasm media management system used for the project and its workflow model will be discussed.

This talk will introduce the Phantasm project and its parent project UIHistories http://uihistoriesproject.chass.illinois.edu/, along with their background and vision. The focus of the talk will be on the information management issues associated with the projects, especially the cataloging and management of the images in the UIPhotos project and the workflow model used to handle the project's very high accretion rates. The scope of the collections and the wide-ranging communities they service present a number of interesting indexing challenges that will be explored in detail in this talk. While complete solutions to many of these challenges remain elusive, the talk will discuss the approaches taken by the project and their outcomes. The talk will conclude by introducing new directions being taken in the project such as the use of GIS technology to integrate high-resolution spatial indexing of content.

KEYWORDS: digital media, digital history, preservation, media delivery platforms, digital access, digitization