Doctoral Workshop in Seattle

This is a three-day workshop in Seattle about the NSF funded program on social computational systems (SoCS). It is actually a PI meeting, but has some sessions to give feedback to doctoral students. Amazingly, it is my first workshop in my academic life, and it has been a really good start.

It's my first time to Seattle. Long before coming to USA, I've learnt the beauty and charm of Seattle. There is nothing thrilling in this city, but its quietness is attractive enough for people to live there for a whole life.

I gave a poster presentation on the first and second day. It's about some preliminary design features to meet challenges of online geodeliberation. It was my first time doing a poster presentation, and it was really challenging. Although poster presentation is more informal than presentation, it doesn't mean it is less challenging. The real purpose of poster presentation is not really report your work, but to rise people's interest in your work, and give a vivid talk about the importance of your work and the challenges you are facing. Also people tend to move around, you need to give them a quick overview of your work, rather than go into too much detail. It may always be frustrating to restart over again as new set of people are joining, so sometimes don't hurry to talk from the very begin. Sometimes you can wait till they ask you some questions and jump directly to that point. Of course you have to keep the balance.

The workshop is a great opportunity for me as I met many professors and students (from University of Washington, MIT, harvad, Ohio State, Oregon State, Virginia Tech, etc. )who are working on social computing, which seems a very promising direction. Many are utilizing social media data (e.g. twitter data) to identify interesting patterns, using data mining techniques. Another popular topic is crowdsourcing, which takes advantage of the power of the public.

As a professor summarized, computer scientists always try to think of solutions to a problem, while social scientists devote themselves to better understanding of the problem, and we working on social computing try to bridge the gap, and always start with people, observe their current practices and identify how their behavior could be improved and facilitated, and design technology to support meet those challenges.

Finally I'd like to put a word I learnt from the workshop:

"I alone can't change the world, but I can throw a stone across water to create many ripples". 

 ― Mother Teresa

GIS as Communication Tool

Previously I viewed GIS as more of a professional tool -- although I tries to turn it into commercial use by common people. GIS was used in the context of one person, one computer. The user take advantage of GIS to make spatial analysis. I believe that is largely why GIS has long been targeted to experts only.

In fact, another big facility of GIS is to serve as a communication tool, among a group of people, or even a whole community, in addressing a spatial issue. The difference between the two uses of GIS is data. In the first case, data has been collected from various sources (often in large amount), and ready to be cleaned and analyzed. In contrast, in the communication context, data is originally stored in user's mind. As he describes his spatial proposition or spatial experience, he either sketches a map or makes reference to a spatial feature. In other word, the map is dynamically created, and is dynamically changing.

The communication problem is actually a problem of information flow. In a situation of two-people dialogue, information flows from one participant's mind to a physical or digital information carrier (often a map in spatial issue), and from the information carrier to the other participant's mind. The process involves the translation from spatial mental model to system conceptual model, and again from conceptual model to mental model. There are two difficulties in the translation: 1) how to minimize the loss of information; 2) how to keep information on the receiver end as consistent as possible as that on the sender end.

The problem becomes even more complex when the social scale becomes larger, to a community, for example. People of various backgrounds are involved in discussion. As they make spatial reasoning and deliberation, they communicate their spatial proposals, assertions, comments, and arguments, which form a large reservoir of information. Furthermore, different pieces of information is highly related to each other. This relation might be linguistic structure, topic embedding, or spatial related. How to keep management of the complex structure of information is worth considering.

According to Walton and Krabbe's typology, human dialogues can be categorized into six: information-seeking, inquiry, persuasion, negotiation, deliberation, and eristic dialogues. Each type of dialogue bears different characteristics, and requires different GIS model.

GIS as a communication tool can be a very promising direction, which involves multiple disciplines like linguistics, social science, cognition science, HCI, etc.