TED Talk: The web's secret storyies by Jonathan Harris

Last Lecture by Randy Pausch: Achieving Your Childhood Dreams

ONE.org

Hmm... Shit.

Dokdo, the Korean territory

In Korea, another candle is burned today. Dokdo or Tokto, which was never been called as Liancourt Rocks since its appearance in history, is very small islands in the East Sea (Eastern Sea of Korea). They are very small in size, but very important for our pride.
I'll never forget the fucking Japan, fuck!.

Skate where the puck's going, not where it's been

by Wayne Douglas Gretzky

Full of Sadness

...

Shame of my Life

I don't think he can only imagine, but he realized all.

New Place New Life but Same Salvation

Since yesterday and until someday, I have been breathing this place, Jeju -- the largest island in Korea. Specifically, I'm working for Daum Communications, and my workplace is GMC (Global Media Center) in Jeju. Here is a fantastic place to work, as well as to refresh my life. I don't like to live in JAM of crowd. I belong to Data Mining Team, of which mission is to improve search quality and related works. Particularly, I'll be in filtering and recommendation of information, which I have been working and intereted in for a while, and until now. I made somehow a big plan, but I ain't sure how much I will achieve it. But, I'll try for it. Anyway even though I started a different, I mean new, life in new place, as repeated Jeju, the gospel is that the salvation belongs to Jesus Christ.

BusinessWeek: Generation MySpace Is Getting Fed Up

BusinessWeek posted an article about escapers from (seemingly free) social networking sites. See the article here.

One of the 2008 Tech Predictions by PCWorld was the collapse of social networking because of no source of solid earning-money. SNS has no way except to show advertisement for survival. I cannot deny that I was one of mean (one-dimensional) persons to think how to effectively (that is, not annoying readers) attach ad's with original content. But the BW article makes me think again. It is obvious that the situation is becoming worse. Dual service - free posting with advertisement or paid posting with ad-free.

Gossip: Wisdom Era: The Futures after a Future

It was a natural progress from data (data mining) to knowledge (knowledge discovery) through information (information retrieval and extraction). Many people, indeed most of so-called gurus, talk that the future is an era of knowledge. Whoever has knowledge will be the only survivor. Who has the cutting-edge knowledge (a funny thing is that knowledge = Know + L + Edge) will conquer the survivors. This may be true, but we must also prepare the futures of the future (knowledge era?). The only word I can think is 'Wisdom' -- beyond knowledge. For a while, someone who knows something will earn money. At the next, someone who knows everything will earn money. At the end, someone who not only knows everything but also integrates them will earn money. Wisdom is not just knowing everything nor simply merging them. In addition, everything does not mean just technologies or economics or other individual stuffs. It means technology, human, God, culture, economics, environment, and so-called everything in the world. Someone suggested that every enterprise must has a position called CKO (Chief Knowledge Officer) as well as CEO, CFO, CIO, COO, and so forth. By the same sense, future enterprise must have a position called CWD (Chief Wisdom Officer).

Presentation Delivery like Steve Jobs Done

BusinessWeek published an article that helps you improve your presentation as Steve Jobs, Apple CEO, does. Followings are ten general and well-known, but often ignorable, principles.
1. Set the theme
2. Demonstrate enthusiasm
3. Provide an outline
4. Make numbers meaningful
5. Try for an unforgettable moment
6. Create visual slides
7. Give 'em a show
8. Don't sweat the small stuff
9. Sell the benefit
10. Rehearse, rehearse, rehearse

View Details via BusinessWeek:: Deliver a Presentation like Steve Jobs
and other related articles (written by the same author Carmine Gallo)
- BusienssWeek::Steve Jobs' Greatest Presentation
- BusinessWeek::How to Wow 'Em Like Steve Jobs

Machines Become Smarter than Human?

One of common thoughts from many gurus -- both optimists and pessimists -- on futures study is that machines will eventually become smarter than human being. Yes, machines will be smarter than human, but will not be smarter than machines themselves. I am always suprised by their -- futhre gurus -- surrealisitc thoughts but also disappointed by their limited thoughts... like this. The only poor who makes something better than itself is human being. We have been trying to develop a smarter and smarter machine -- robot -- both in kenetics/mechanics and in (artificial) intelligence. The future pessimist predicts the smarter machines will develop another much smarter machines, and then the smart'est' machines will govern human beings. This is the future of pessimistic human beings' society. OK, this is possible. But, the thing they have missed is that if a machine is, indeed, smart and intelligent it will not develop more intelligent machines than itself, because it [the smart machine] knows the smarter machines will destroy/govern the "old-fashioned" creator machine itself by the same course of the tech pessimists. Therefore the smartest machine is no more intelligent than human beings. But, as I metioned, human beings are poor enough to develop new gods, the new conqueror, -- intelligent machines.
I'm not ignoring the rapid advancement of computers -- both hardware and software, particularly artificial intelligence.

Discussions of Tech Predictions in 2008

No comments on my own tech prediction for 2008.
Visit http://www.mahalo.com/2008_Technology_Predictions to inquire a list of 2008 predictions.

Ten Problems and Ten Algorithms in Data Mining

This posting is based on a public mail disseminated by Xindong Wu, who has gathers reports from data mining and machine learning reseacher groups over the world during 2005 and 2006 the IEEE International Conference on Data Mining (ICDM).

10 Challenging Problems in Data Mining
- Developing a Unifying Theory of Data Mining
- Scaling Up for High Dimensional Data and High Speed Data Streams
- Mining Sequence Data and Time Series Data
- Mining Complex Knowledge from Complex Data
- Data Mining in a Network Setting
- Distributed Data Mining and Mining Multi-agent Data
- Data Mining for Biological and Environmental Problems
- Data-Mining-Process Related Problems
- Security, Privacy and Data Integrity
- Dealing with Non-static, Unbalanced and Cost-sensitive Data

10 Most Influential Algorithms in Data Mining
- C4.5: Decision Tree
- k-Means: A typical clustering algorithm
- SVM: Support Vector Machine
- Apriori: (No Knowledge)
- EM: Expectation-Maximization
- PageRank: Google's search ranking algorithm
- AdaBoost: (No Knowledge)
- kNN: k-Nearest Neighbors
- Naive Bayes
- CART: Classification And Regression Tree

For details of them refer to following URLs:
- http://www.cs.uvm.edu/~icdm/10Problems/index.shtml
- http://www.cs.uvm.edu/~icdm/algorithms/index.shtml

Diary... What's NeXT?

In the last posting, I posted my recent works and thoughts and so on. Last a few days I have felt I encountered a wall that prevents me from keeping going forward. I had many ideas about research works worthy enough to write papers. But, from my naive implementations and experiments, they are proven not worthy spending time on them. Such dishearted feelings may be a natural process of my life, and of yours as well. I'm accepting that, but also I'm tired. I'm exhausted. Research works are the only things I can do now and what I'm best. They are only things that make me feel alive. I have spent much time in a variety of topics, at least, after my MS entrance. In the undergraduate course, I'm a person having no thinking. I also critize them for the same reason. At first, as my lab -- manufacturing systems integration -- I studied about manufacturing systems including the lowest level of NC code to somehow high level automation and integration, and intelligence, and so forth. At that time, I also spent much time on PLS stuff for LG Micron. We hated that project but it gave us many things even though we could recognize them. After that, we focused on e-business stuffs. The main reason was we participated in related projects, particularly KorBIT. We thought Interoperability would continue to feed us. I know that even now interoperability is still far away from us. Before my US life, I was novice in research, indeed. The life in US made me changed, even though you may not recognize my changes. First, I have an eye to see something. Second, I have brain to think something different, not just new. I like 'different', but hate 'innovative', even though my mouth always keeps the word. The most important is I have a heart as others have. Yes, I also have a heart to feel. I'm a person who has no empathy with others, actually. Still, I'm such a person. But, I can say that I also have a heart. I was a narrow-sight person. During and after US life, I was focused on XML technology, especially on similarity between XML. I'm still working hard on it, but no visible outputs, because I am just thinking it. If I have an environment only to focus on it, I think I can change the ENTIRE world. However, I am not in such a good circumstance. I could graduate with the XML similarity. The dissertation is not good. It cannot satisfy me. I am never satisfied with it. If I can go back to 4 or 5 years, I can make a surprising thing. After graduation, so-called Dr. Jeong, I was concentrated on data mining. For the long research records on XML, at first, I also studied data mining for XML. It was a natural progress. At that time, I knew semi-supervised learning and kernel methods. They fed me for a while. Perhaps, until now. From text mining, especially VSM, I found the possibility to apply it to recommender systems. But, there were already many works done in recommender systems, especially collaborative filtering. However, I could recently do something about collaborative filtering. That makes me also feel alive, but also discouraged me because no improvement is expected. But, I'm still thinking about collaborative filtering, because no one is surprisingly good enought to meet my expectation. Manufacturing system, PLS and statistical methods, B2B and interoperability and testbed, XML and similarity, semi-supervised and kernel, and collaborative filtering. This is the full track of mine. At thirty, I need another weapon to spend my energy. What's next? is the only question I always try to answer. But every time, silence only. Thank God for your silence. I always want to hear your voice, see your glory. But you never showed me through visible things, spoke me though audiable things. You always implant on my heart that I found during my lowest life time. I wanted to be honest to my life to my research works to my neighbor and to You. But, I always failed. So, I'm trying again. ... Time waits for no one, but You do.

mySelf @ Pre-SenT

It's long time since I posted. The time on me was to focus on technical stuffs, so it was hard to post my ideas and thoughts on this blog. Currently issues I'm concentrating on are about Christianity, which is my primary for ever and ever, and the others are technical stuffs such collaborative filtering, which is a well-known and most famous algorithm for recommender systems, similarity measures between XML documents same as last years, and social networking as well. I'm not focusing on the technical aspects of the social networking such as AJAX or other programming stuffs, but on general aspects of it to the surrounding society and especially to economics. Particularly, on how to earn money from social networking. Today's newspaper gave an interesting article about what will happen in 2008. The most top issue is the fall of Web 2.0 --- actually Web 2.o companies due to no source to make profit from their social networking sites. The sites are attractive to many users but they never spend money for them. This is the dilemma the Web 2.0 companies face. Indeed, the concept of Web 2.0 will last for a while -- participation of volunteers. ... Collaborative filtering is the hottest issue I'm currently involving to do something -- particulary preparing scientific papers --- actually I'm ready to submit at least two written papers now, and also preparing other two or more papers on it. The next step may be to appy the algorithms I have developed on real applications. To do this, I need a big-spending project and supports from others (researchers and developers and espeically evangelitic friends), or getting into a new world -- professional job. One of the most annoying things at this time is what I will do next months or years. I msut be now an eager job hunter, but I'm tired and everything does not work as I expected. I want to be MD. USA nearby friends in Christ. ... It's terrorfying me... What's next to me?