Friday, December 13, 2002

Graph

My.network

[ click for colored full size image ]

This Bud Network represents the relationship among people that are close to me. Different color represents different type of social relations, such as drinking buddy or band buddy. Through the graph, the relationship among my friends are mapped out. In this graph, the type of relationship that each person is having with one another can be easily idenfity. Some people have more social life, so their links will be more. In this case, I am the person who have the most links, because this map is constructed primarily based on my social relation. In averge, each person will obtain similar amount of links, and if any of these people want to find another person in my bud networ, he/she can find a way to relate to that particular person.



Google.network

[ click for colored full size image ]

Close Network and Open Network

Using Google as start point
Category: Game > Video Game
Selected Node: IGN.com

Close Network
IGN is a collaborative website which contains entertainment information regarding game and video. It acts like a news website. IGN contains massive data and a huge internal network; it has several sub pages, such as games.ign.com and entertainment.ign.com. The information in sub-page cannot be linked to outside resource. The only links out are to gamespot.com, which is similar to games.ign.com and an online store, gamestop.com. IGN serves as a close network because it keeps all its information to itself. IGN is determined to create an organizational community.

Open Network
One the other hand, Gamespot.com is an open network. It functions as a game news website, but the user can link to specific external information source. For example, the user can link to a specific game’s original official website, such Sony.com, nintendo.com (this function is not allowed in IGN). Also Gamespot.com has a close relation with CNET, they exchange link and share some related news. Regarding the out link to similar website, the user can be direct to many other websites, such as IGN.com, GameZone.com, Gamer.com, through GameSpot.com. Gamespot.com is determined to create an open online gaming community.

Pathway
I started off using Google to search for video game website. I chose IGN.com as the starting node. I obtained IGN.com’s internal structure by surfing around and using the TouchGraph tool. I found a link to link to Gampspot.com. And from there I link to several others website.

Analysis
Both IGN and Gamespot.com have a great amount of gaming information. Being in the game industry field, Gamespot.com contributes more on building the network community by being an open network than IGN.com. Traveling in IGN.com, the user will find a lot of information but he/she can hardly get some related information. However, in Gamespot.com, the user can travel around in the gaming industry.

Wednesday, November 13, 2002

POP Talk vol.x

When pop music meets complex system

Music is a part of human culture. Throughout history, civilization, music has a profound influence on the human society; it has different forms and it has different meaning from culture to culture. As globalization emerging, there is lots of musical mergence, and birth of new music. What people called pop music today, is merely a collection of music style that most of the people accept (that¡¦s why dark metal can¡¦t be accepted as mainstream, so it stays underground as a sub-culture). Looking at the whole music culture, we can undoubtedly find numerous links, since all kinds of style have some sort of connection between one another. This category is rather broad. Focusing down to the pop music, we can see that it is a complex system itself, and it demonstrates a more clear relation among all the elements within. In a way we can use the title ¡§pop music¡¨ as a framework to map out the links that all the styles share. There is actually a number of ways we can define the network within the framework. First, pop music embraces a variety of music styles. If we categorize them, interestingly, we can find a few huge hubs; for example, rock music is always popular and it can link to rap or ballad culture. These hubs gather characteristic from all kinds of music style, and getting more and more popular because they embrace a wild range of techniques. Also, we define the network by a certain style of artists. Say¡Kbackstreet boys, their songs include a variety of styles, such as cheesy crystal love song, hard rock, or even Spanish style. And people use these artists to define music era. To sum up, the purpose of this session basically is to relate the course material to my own interest. Since we are talking about culture and complex system, music plays a major part in this field. There is no doubt that music has a great amount of culture value; even its contemporary form, which is the pop music, has its own culture implication. Pop music is what we hear each day; in a way it defines our generation (contemporarily of course). This issue is actually pretty deep, definitely worth investigating.
Blogging Experience

Looking back at the blogging I have done in the past 5 weeks, I think there is a clear demonstration of steps of learning, like some sort of learning milestones. Essentially my insights are revolving around the same topic, the Internet; however, each week, I tackled different perspectives, as I dig deeper into the book, Linked. First I started off figuring out the random network, and shift to the scale-free network applying on the Internet technology. Then I talked about psychological connections between the user and the technology, like how the Internet becomes a dialog between users on different ends. It is interesting to see how knowledge is developed based on the precedent ones. In a way, I think the blogging experience is a complex system. Each issue, such as n degree of separation, power law, has its own meaning of existence, and by putting the ideas together, they achieve a greater meanings, ranging from something big as rocket science to something as tiny as micro-biology. Same elements are used to explain various concepts or phenomenon. It is interesting to see how the physical network, such as the airplane route, maps to the virtual network, such as the Internet. In the past weeks, through all kinds of readings and researches, I am able to see it myself how the individual concepts here and there are used to develop a greater concept. I think the book, Linked, demonstrates this idea as well, seeing how the author goes from the Internet to how our body cell works, which is rather difficult to me since everything looks like a biological genre to me. But anyway, the point is, everything is linked, though sometimes the link can¡¦t be easily seen. The endless postings talking about ¡§linking¡¨ actually demonstrate how everything is a part of a big complex network (seeing how each my articles link to each other). Yes. Even this statement makes everything even more complicated than it already is. After all, anything can always be complicated if we want to apply a certain network onto them ¡K no? (These types of thinking really drive people nuts¡K)

Tuesday, November 12, 2002

POP Talk vol.5
Pop n' Life



Here we are
Doesn't really matter we feel like or not
We are surrounded by it
Soon we will accept it
Then we will become a part of it
like each one of us share a piece of it

Friday, November 01, 2002

diffusion in a complex network, the Internet

The other day I saw an interesting incident. There was this video clip on the Internet; the link was sent to me by one of my friends. In this video clip, there were 2 Korean girls singing Karaoke, singing a fast and interesting song. There was this over-weighted girl singing and dancing. It was funny (not that I despite her kind of way, it was just funny with all these dancing and funny song). I am guessing she just uploaded this clip to show her friends (because if it was me, it would be the only reason I would upload this clip). Somehow it ended up spreading nation-wide, considering I don¡¦t know the girl, and I am not familiar with any Korean. The funny thing is, when I tried to see the clip again 2 days after, the clip was gone. I assume that the girl somehow became aware of the situation.

This example in a way demonstrates the diffusion in a complex network. The video clip was originally uploaded to a website. Somehow the link started to spread out, through emails, ICQ¡Ketc. Aided by the fact that people like to see freakish funny stuff on the Internet, the link spread out like crazy. I don¡¦t know the girl, I don¡¦t live anywhere near Korean, I hardly know any Korean, but I saw the clip and I forwarded to my friends. The most interesting thing in this incident is that since the Internet is so interactive, interlinked, somehow the original creator knows that the clip is being seen by perhaps thousand of people, and she decided to take it off.

Diffusion in a complex network is an interesting thing, especially with the Internet, where the information can spread out more thorough, since the Interest is popular and highly accessible. Once the information being spread out on the Internet, major hubs will be created. Informative sites such as BBS and news site become influential hubs. Usually the information that is passed around on the Internet has a relatively high spreading rate; people pass it around because it is interesting, like jokes or funny pictures. How popular this information is, is determined by the spreading rate. If the information is really worth passing, the spreading rate will be over critical threshold, and this information can reach numerous viewers. In this case of dancing girl, I believe the spreading rate is way over the critical threshold.
Tolerance against Failures and Errors

When building a network, any network, it is important to consider its tolerance against failures and errors. Higher degree the tolerance is, more stable the network will be. The term robustness is used to describe a thing's strength or ability to survive/adapt under extreme condition or change in environment. Robustness of a network describes this network's stability and its strength against malfunction when local failures occur. Usually a highly interconnected complex network will have a higher tolerance against failure. A networks robustness depends on its structure.

When local nodes start breaking down, isolated fragment will start appearing in a network. If the number of the failing local nodes is relatively small, although the behavior of the network might shift a little bit, there will only be an insignificant impact on the network's integrity, at least the system can still be working. When will the failure nodes start damaging the system? When one determines a networks degree of tolerance against failure, one will look for a threshold that distinguishes a working network and a collapsing network. In a random network, the threshold is obvious. When the number of failing nodes reaches a certain amount, the network system will start collapsing. On the other hand, in a scale-free network, the threshold is relatively small, where the degree exponent is smaller or equal to 3. In the other word, if the node that ceases functioning occurs randomly, the chances of having a total failure is almost zero.

A scale-free network might be stable in a normal daily use where errors might randomly occur; however, it is weak against the intentional sabotage. The use of major hubs decreases the systems chance of breaking down, when the local nodes fail, but at the same time the major hubs become an obvious target, if someone wants to bring the system down. A simultaneous removal of major hubs will have dramatic effect on the whole system; it will collapse in no time. Although it requires a deliberate plan and techniques to do so, a scale-free network can hardly prevent this kind of attack.

Another issue of the failure of a scale-free network is the cascading failure. While it makes the system stable by shifting the load of a failing node to others, when the neighbor nodes cant afford the extra load, nodes will keep failing one after another. In a real life example, in TechBC, we emphasize the importance of team work. Say there are 6 people in a major project and each person is assigned an important task in different area. If one person fails to finish the part (lets say that he suddenly decides to drop out the course), his/her share of work load has to be transferred to other members, mostly likely the person who can handle the task in this particular area that the dropping out person was assigned to do. What if this person collapse because he can't handle a double work load, then some other members might need to pick up the piece? Since the tasks were equally distributed in the first place, a sudden change in task arrangement might drag the whole project down. And this kind of thing happens a lot (in my experience)

Tuesday, October 29, 2002

POP Talk vol.4
Mainstream




Pop culture defines mainstream. A style/concept that is more popular and accepted by most people will most likely become mainstream. Styles are everywhere; music alone contains numerous styles from different cultures. With all these styles swimming around in the tank of music industry, once in a while a style will rise up to the surface of the pop culture and become a part of the mainstream. Sometimes these styles exist in a form of sub-culture. For example, rave/techno music is originally the essence of the rave culture; it is a part of this culture. As more and more people accept the techno music, it evolves to a pop culture itself. Artists like, Moby, Madonna, Kylie Minogue, have a significant influence in the field of techno music.

It is hard to describe mainstream because it is dynamic. What popular today might not be popular tomorrow. Artists work hard day and night, album releases one after another, just to maintain the popularity. There is no doubt that this particular trait allows music industry always be fresh, which is good for the audience¡¦s senses; we always want new thing. But one thing we really need to consider, and that is the quality of the music. Boy band emerges one after another. They are popular alright, but is it just another way for the music company to make profit?..

Sunday, October 27, 2002

Search Engine n' scale free topology

Scale-free model represents a network where the links are unevenly distributed and all the nodes coexist with one another. The topology stands when the new comer assigns links its links to a certain number of any nodes. However, when it comes to the Internet, the fitness of one node needs to be considered. Not any of the nodes might attract the links; the node that has a higher fitness level might attract more links. The node that acquires the most links becomes the fittest one.

This idea is indeed important in the web; this way the construction of the Internet will be more logical for people. Search engine is the best example. Search engine is no doubt one of the most important components of the Internet. It allows people to obtain information more efficiently, because it serves as a huge hub which contains and categorizes numerous links. A search engine constructs in an organized fashion that the viewer can easily understand; it becomes some sort of translator allowing the viewer to visualize the links. Google, for example, is one of the strongest search engines in the world. Its database is widely used by other search engine nationwide, such as yahoo. There is no surprise why Google has such a high popularity, because it makes people's life a lot easier.

Google Network

Thursday, October 24, 2002

Power Law

In a power law, the nth item generally has a size or frequency of 1/n of the largest -- the tenth largest city is about one-tenth the size of the largest, and this emergent higher order organization is a feature of many human systems
--In Defense of Cities

While scale-free network explains how there are rules behind complexity, power law is the center guide of the functionality of a scale-free network. As mentioned previously, the power law distribution indicates that in the most of the real network, the majority of the nodes have only a few links; they connect to a few big hubs, and small nodes and hubs coexist to form an active network. Unlike most of the nature's behaviors, power law is against to the idea of decentralization where most nodes are equally distributed. The graph of the power law distribution has a tenancy of gradual decay as it moves towards the end. Characterized by a unique exponent, power law shows the relationship between the high end and low end as a whole. Under power law, since the distribution of the links among nodes are not even; they are formed in a relation of hub and nodes, scale-free network can keep expanding and grow as a whole.
Scale-free network

In a network system where all the nodes do not share the same characteristic, the distribution of the links will not be not even. All the individual nodes exist in a context where they are a part of some larger event. In a scale free network, the behavior of all the nodes should be study as unison, since the links among nodes are governed by power law; most of the nodes have few links and they are held together by a large hub.


Scale-free network has a unique nature, earlier a node appears in the network, more likely it can obtain more links. In the scale-free model governed by growth and preferential attachment laws, each new node is given 2 links that it can attach itself to other nodes. The node that shows up earlier in the system will have more chances to be connected by the latecomer. The concept is not hard to understand. Say if I were to decide which website I want to put my company¡¦s link to, I would for sure go for the most popular one. Since there are more people surfing that popular site, the chance that my company¡¦ link be spotted by those surfers will increase. From the business point of view, selecting a major hub allows my individual node be more accessible by the public (considering the public will access the major hub more than the minor ones).


While the popular sites keep growing, and become the center of a cluster, it is very likely that they become the target of the network hazard. Due to the nature of scale-free network, small sites are connected to a major websites. As the connection being established, computer virus might seize this behavior and use email as route to spread out. A well-established network becomes a mean for virus distribution; this is why a single virus can have a global impact and cause billions of damage.