Quantifying social vs. antisocial behavior in email networks

Luiz H. Gomes,Luis M. A. Bettencourt,Virgilio A. F. Almeida,Jussara M. Almeida,Fernando D. O. Castro

DOI: https://doi.org/10.48550/arXiv.physics/0601141

2006-11-28

Abstract:Email graphs have been used to illustrate general properties of social networks of communication and collaboration. However, increasingly, the majority of email traffic reflects opportunistic, rather than symbiotic social relations. Here we use e-mail data drawn from a large university to construct directed graphs of email exchange that quantify the differences between social and antisocial behaviors in networks of communication. We show that while structural characteristics typical of other social networks are shared to a large extent by the legitimate component they are not characteristic of antisocial traffic. Interestingly, opportunistic patterns of behavior do create nontrivial graphs with certain general characteristics that we identify. To complement the graph analysis, which suffers from incomplete knowledge of users external to the domain, we study temporal patterns of communication to show that the dynamical properties of email traffic can, in principle, distinguish different types of social relations.

Physics and Society

What problem does this paper attempt to address?

Based on the provided text content, the problems that this paper attempts to solve can be summarized in the following ways:

### Research Background and Problem Description

This paper mainly studies the degree distribution and clustering coefficient in social networks and anti - social networks. Specifically, the author focuses on:

1. **Degree Distribution Characteristics of Social Networks and Anti - Social Networks**:
- By analyzing the degree distribution of nodes in social networks and anti - social networks, it is studied whether these networks follow a power - law distribution, that is, whether there are a small number of high - degree - connected nodes (hubs).
- The paper obtains the power - law exponents of different networks by fitting data and evaluates the goodness of fit.

2. **Clustering Characteristics of Social Networks and Anti - Social Networks**:
- Analyzes the distribution of clustering coefficients in social networks and anti - social networks and explores the local structural characteristics within these networks.
- Compares the differences between actual networks and random networks to understand the non - random components in actual networks.

### Specific Problems

- **Power - Law Characteristics of Degree Distribution**: The paper attempts to verify whether the degree distributions of social networks and anti - social networks conform to the power - law distribution and determine the specific power - law exponents. For example, for social networks, the power - law exponent is \( \alpha = 1.82 \), while for anti - social networks, the power - law exponent is \( \alpha = 2.03 \).

- **Distribution of Clustering Coefficients**: Studies the distribution of clustering coefficients of nodes in social networks and anti - social networks to understand the local connection patterns in these networks. By comparing the distribution of clustering coefficients between actual networks and random networks, it reveals the community structures or other complex characteristics existing in actual networks.

### Formula Representation

1. **Power - Law Distribution Formula**:
\[
P(k)\propto k^{-\alpha}
\]
where \( P(k) \) represents the probability of a node with degree \( k \) appearing, and \( \alpha \) is the power - law exponent.

2. **Log - Linear Regression Model**:
\[
\log(y)=\log(b)+\alpha\log(x)
\]
where \( y \) and \( x \) are the dependent variable and the independent variable respectively, \( b \) is the constant term, and \( \alpha \) is the slope (power - law exponent).

3. **Clustering Coefficient**:
The clustering coefficient \( C_i \) is defined as the ratio of the number of triangles formed among the neighbors of node \( i \) to the maximum possible number of triangles. For the entire network, the clustering coefficient \( C \) can be defined as the average clustering coefficient of all nodes.

### Conclusion

Through the above analysis, the paper aims to reveal the similarities and differences between social networks and anti - social networks in terms of degree distribution and clustering coefficient, thereby providing a theoretical basis for understanding the structural characteristics of these networks. This is helpful for further studying the evolution mechanisms, information dissemination characteristics, and potential application scenarios of networks.