IST590 Fall08

I have read the one by Marie desJardins, who actually works on intelligent agents, and I have cited her papers in my own.

Interestingly, two of the three articles were written by computer scientists. Maybe computer scientists spend too much time in the laboratory and play with algorithms. Thus they tend to ignore some other important factors that lead to success. From this perspective, IST is different from computer science departments, partly because IST is more eager to make its student successful, and IST does give its students more diverse advice on how to succeed. I am happy to see that many of the points in those three readings have been mentioned by my peers, teachers, collaborators and advisors in various occasions during my stay at IST.

The basic ideas about how to survive or flourish in graduate schools remind me what Dr.Yen said at the IST career workshop: consider yourself as a scholar even though you are only a graduate student now.  Basically, pursuing PhD is the starting point for a career in research. We should carefully plan it, work hard for it and cultivate other non-academic skills and networks in order to succeed in it.

I am now an active member of the Penn State Chinese Friendship Association. I maintain the website for the organization and have participated in the organization of events such as the new Chinese students welcome reception, the Run for Beijing Olympics, the Mid-autumn Festival Party and the Chinese Culture Festival. The involvement in this association provides me an opportunity to know the Chinese student community (there are about 800 Chinese students in the UP campus) in Penn State and to help people from other countries to learn more about China. I made a lot of friends who share the same hobby, the same hometown or the same former schools with me.

Interestingly, one of my photos at the Chinese Culture Festival has been uploaded to the organization's online album and one of my college friends, who was just about to come to Penn State as a visiting student but didn't know that I am in Penn State, saw that photo and thus re-connected with me.

I play badminton, soccer and table tennis with my friends occasionally and may join one or two university-wide club(s) in the future.

When I was asked to write this blog post, the first name popped up is Kathleen Carley. Although Dr.Tapia stole my thunder by mentioning her as an example at the class, I still decided to write something about her, because I have been hearing her name and citing her work for quite a while and searching for 'Carley' in my EndNote will return 29 papers.

Dr. Carley spent her years as a student at Boston. She received S.B. degrees in Economics and Political Science from MIT and her Ph.D. was from Harvard Sociology. Although Dr.Tapia considered her as a sociologist, I believe the training in Economics at MIT gave her a different perspective and skill set. She is definitely an outlier among sociologists. I would rather call her an interdisciplinary researcher with a very diverse academic background. Sociology might just be one of her 'gas stops' along her way in interdisciplinary research. Sorry Andrea :)

She is now a professor of Computation, Organization and Society at the School of Computer Science, Carnegie Mellon University.  Due to the interdisciplinary nature of her research, she also holds join positions at the Institute for Software Research, Social and Decision Science, Public Policy and Management, Industrial Administration and Engineering and Public Policy at Carnegie Mellon.

Her research combines cognitive science, social networks and computer science to address complex social and organizational problems. Her specific research areas are dynamic network analysis, computational social and organization theory, adaptation and evolution, text mining, and the impact of telecommunication technologies and policy on communication, information diffusion, disease contagion and response within and among groups particularly in disaster or crisis situations.

So where and why is she famous?

I would say she is most famous in the community of computational social and organizational scientists. From my experience, most active researchers in this area know of her and a lot of them have worked with her. Why? She is one of the creators and pioneers of this emerging area, although the area is still under development. She is the missionary who introduced new techniques to tackle problems that cannot be solved with traditional approaches. She created and directed a research center that has about 40 staff and students. She founded a professional association (North American Association of Computational Social and Organizational Science) and played important roles in the development of the journal of Computational Mathematical and Organizational Science.

She is also known to some computer scientists and has published in AI communities. In my opinion, her major contribution to AI, especially distributed AI (or multi-agent AI), is bringing new real-world problems that haven't been considered by computer scientists and thus expanding the application domain of computer science research. More importantly, when solving new problems, she also discovered new techniques and contributed back to computer science research, such as data mining and network analysis. To my knowledge, she is among the few researchers who can connect the computational social and organizational science community and the AI community. That is probably also why she's now at a computer science school (or maybe it was her position in a computer science school that enable her to connect with computer scientists).

I am glad she will come to IST and am looking forward to her talk.

Hey presidential candidates, be nice to techies. Give them tax breaks and spend the $700b package on Silicon Valley. Otherwise, who knows what will happen to your votes:)

Suggested reading for political figures who are campaigning:
The IEEE Spectrum article "Want to Hack an E-Voting Machine in 7 Minutes?" (Full article).

A few interesting sentences from "Mining for data" from CBC Canada (Full article):
-------------------------
Business writer Stephen Baker says in an interview that our casual disclosure of personal data in our daily lives provides grist for computer scientists, mathematicians, programmers, and others who are mining the data to make sense of it. These people, which Baker dubs the Numerati, are modeling and predicting our behavior, and the applications borne from their efforts are dramatically influencing our lives. Baker argues, for instance, that the Numerati are to a certain degree accountable for the global financial crisis because banks' investment decisions are based on algorithms that the Numerati designed according to their understanding of risks, or lack thereof. "It's important to understand that you can have the best math in the world, but if you don't understand human behavior, then you cannot calculate risk when it comes to market behavior," Baker asserts.
-------------------------
Maybe it is promising to integrate human factors into the math-heavy data mining of nowadays.

From CNET.com

The U.S. intelligence community is concerned that terrorists might use micro-blogging tool Twitter to coordinate attacks, according to a purported draft Army intelligence report posted on the Web.
(Full article)

I have listed many publication outlets in my previous blog introducing research communities. Here, I picked a few that I am currently aiming at or particularly interested in:

1.  AAMAS conference: the top conference dedicated to intelligent agent research. This annual conference was first held in 1997 as the Autonomous Agent Conference. Almost all renowned agent researchers go to this conference. The recent acceptance rate is about 23%.

2. AAAI conference: one of the top AI conferences in the U.S. The acceptance rates span from 17% to 30% in its 29 years of history.

3. IJCAI conference: one of the top international AI conferences. It is held every two years starting from 1969. The 2009 conference will be at Pasadena, Calif. and the submission deadline is Jan 7, 2009.

4. The ACM Transactions on Modeling and Computer Simulation: it was first published in 1991 and has four issues per year. Each issue has 3-5 papers.

5. The Computational and Mathematical Organization Theory journal: this is an interdisciplinary journal published by Springer. It started in 1995 and publishes 12-20 papers every year.

Due to the interdisciplinary nature of my current research, it is difficult to find a single research community that defines my interest. Actually, I may have to adjust the focus of my research to suit a research community.

At this moment, the following two research communities draw my attention:

First, the community of computer science researchers focusing on artificial intelligence (AI), especially intelligent agents and multi-agent systems. Most researchers in this community are associated with ACM and IEEE. AAAI is the leading organization for American AI researchers.

For AI and intelligent agent researchers, there are two types of publication venues: one type is dedicated AI venues, such as AAAI's annual conference, IJCAI, AAMAS, the International Journal of Intelligent Systems and the Artificial Intelligent Journal; the other type is not dedicated to AI but often has publications that apply AI techniques to specific problems. For example, AI researchers have been active in    ACM E-commerce conference (EC), Knowledge Discovery and Data Mining Conference (KDD), International Semantic Web Conference, ACM Transaction on Modeling and Computer Simulation, etc.

Second, the community of scientists who use agent-based computational approach to study social, organizational, business, epidemiologic or geographical problems. Research in this community can be found in venues such as the World Congress of Social Simulations, the NAACSOS conference, Computational and Mathematical Organizational Theory, Advances in Complex Systems, etc. As the computational part is often considered as a research method, researchers also tend to publish papers in outlets that focus on their own research domain, such as Organization Science, Management Science, MIS Quarterly, American sociological review, International Journal of Geographical Information Systems, etc.

The boundary of the two communities is obvious most of the time and few researchers are active in both communities. Therefore, I may have to choose one as the major community that I would like to belong to. Depending on the topic on my dissertation, I may have to explore other research communities.

For this blog post, I talked to Shizhuo Zhu, who is the most senior student at the Intelligent Agent Lab. He has been at IST since 2004 and is now preparing for the dissertation defense.

His dissertation topic is 'Hypothesis-Driven Story Building: A Framework for Supporting Decision-Making as Partial Information Arrives Over Time'. Basically it is about studying the characteristics and challenges of decision-making when information is incomplete and changing over time, and build a system to support human decision-making in such environments. It will be tested and evaluated with in medical diagnostic decision-making settings.

Based on his interests, he published most of his papers at two types of venues. First, the computer science-oriented AI and agent research community, such as International Journal of Intelligent Systems and the Intelligent Agent Technology conference. Second, medical informatics venues, such as the Conference on Artificial Intelligence in Medicine and the American Medical Informatics Association conference.

Several examples of his publication include:

Shizhuo Zhu, Joanna Abraham, Sharoda A. Paul, Madhu Reddy, John Yen, Mark Pfaff, and Christopher DeFlitch, "RCAST‐MED: Applying Intelligent Agents to Support Emergency Medical Decision Making Teams". The 11th Conference on Artificial Intelligence in Medicine (AIME'07), Amsterdam, The Netherlands, July 09‐11, 2007.

P.-C. Chen, X. Fan, S. Zhu, and J. Yen. Boosting-based learning agents for experience classification. In Proceedings of the  2006 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, p. 385-388, 2006.

Shizhuo Zhu. Story Building and Its Qualitative Identification and Quantitative Measuring of Hypotheses. NESCAI'06, April 28-29, 2006.

Although he is an IST student, he sees himself mainly as a technical person. He would like to build technical systems either theoretically or practically that may change people's life style.

His research is similar to mine in that both of us are using intelligent agent technology. However, his research is more like traditional AI (decision-making with Bayesian networks). If we use the IST jargon, his is on the I-T side. Mine might be more interdisciplinary and lies on the P-T side, although incorporating I is also possible in the future.

From CNET.com

A scientific paper detailing the security flaws in the Mifare Classic wireless smart card chip used in transit systems around the world is being published by the Radboud University Nijmegen. And a researcher at Humboldt University in Berlin has published a full implementation of the algorithm (PDF).

"Combining these two pieces of information, attacks can now be implemented by anyone," RFID researcher Karsten Nohl told CNET News. "All it takes is a $100 (card) reader and a little software."

(Full article)