Networks are made up of dots and edges, but when we give social attributes to the Internet, we find that social networks have unique elements that can be incorporated into the current network.
1. Homogeneous phenomenaPeople who connect with each other in social networks tend to be similar.The formation of an edge in a social network is due to the more connections between nodes, or the same friends, the other is that we often meet in the same place, the first relationship can be seen through the edge of the potential edge of the possibility, but for the second, we can not be seen from the picture. (maybe visualization can, but calculate?) In short, the external environment for the formation of the edge of the problem is worth studying.Quantitative homogeneity: If we want to study a thing, first of all to quantify, quantify the data can be calculated, it is much easier. Homogeneity test: If the percentage of transgender change is significantly lower than 2PQ, there is a homogenous sign. What's that about? In the natural state, assuming that there are two camps in the social network, the proportion is p,q, then according to the random point of view, the p of the camp all the links should occupy all the p^2 so much, the other camp is q^2, the relationship between the side should be 2PQ.If it is larger or smaller than 2PQ, it becomes homogeneous or inverse homogeneity.(High school biology has a ha. The law of Weinberg seems to have some similarities)
2. Mechanisms behind homogeneous phenomena: choice and social impactThis is a translation of the foreign documents, the words are simply too many, in short: similar, the relationship between social networks has a correlation. Is it because of the similarity that builds social networks or is it similar to building social networks? The former is called the choice, the latter called the social influence.Spit again: The Foreigner's book back and forth to say several times to read all the trouble.
3. AttributionThe networks we discussed earlier can become social networks, which are made up of people and relationships, and with attribution information, we can build a new network. This attribution information is called the Community (foci). So we have a two-part picture, some people, another part of the community, and you can study how the two groups of data relate to each other.Social networks and attribution networks are changing at a time, representing a synergistic evolution. There can be three kinds of relationship, namely three-dimensional closure, community closure, membership closure. The first ternary is three people, the second ternary is a community of two people, the study is two people because of the possibility of forming a relationship in the same community; The third ternary is a community of two people, the study is a person because another person into the community probability.
4. Link formation from online dataThis theory is quantitatively measured from the point of view of data.
5. A spatial model of isolation"Birds of a feather, flock together," that's probably it.The model describes the influence and effect of homogeneity on space isolation.I always feel like it's an optimization. The goal of this optimization is to, from an individual point of view, as many people around them as possible, each step approaching the optimal solution. So we can say that when I use a computer to simulate this process, no matter what strategy is used, once my goal is determined-as much as possible to live with the same people around me, the simulation results are similar. But the constraints of the elastic different results in the final concrete result may be different, but the trend is the same.
the trend of the model is convergence.
The 2015.4.1 network and its existing environment