Abstract:How can we discover and estimate major events in complex social networks? Even with ever enlarging networks and data scale? Event detection and evaluation in social networks provide an effective solution, which has become the critical basis for many real applications, such as crisis management and decision making. The existing event detection methods mainly focus on text analysis that is limited in social media content and graphic feature statistic that needs calculate vast variables. Can we find an efficient way for generalized social networks with limited topology information? In this paper, a novel hybrid quantum swarm intelligence indexing method (HQSII) from the perspective of link prediction is proposed for the first time, which includes an optimal weight algorithm (OWA) and a fluctuation detection algorithm (FDA). The innovations behind HQSII lie in three aspects: (1) The mixed index that can universally describe the social network evolutions is proposed firstly, which explores the cooperation of different independent similarity indexes. OWA is further proposed to determine the optimal mixed index that achieves higher link prediction accuracy and better network evolution description than other independent and mixed similarity indexes. (2) To better avoid the interferences of routine network evolution fluctuations, the otherness of micro node evolutions is considered into link prediction. FDA is further proposed to quantify the abnormal fluctuations caused by events. (3) Based on OWA and FDA, HQSII is proposed for all the generalized social networks, which detects events by discovering abnormal fluctuations and evaluates events by analyzing fluctuation trends. Extensive experiments on theoretical and real-world social networks show that HQSII can accurately detect events and quantitatively evaluate event impacts in social networks with single and multiple events.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to discover and evaluate major events in complex social networks, especially in the case of the increasing scale of data. Specifically, the existing event detection methods mainly focus on text analysis, which has limitations in social media content and graph feature statistics and requires the calculation of a large number of variables. Therefore, the paper proposes a new method based on hybrid link prediction and quantum swarm intelligence (HQSII), aiming to provide an efficient event detection and evaluation method for generalized social networks with limited topological information.
### Background of the Paper and Problem Definition
With the sharp increase in individual interactions of social services (such as Facebook, e - mail, and telephone) worldwide, social networks have become a research hotspot in many contexts, from uncovering criminal activities to finding friends, from community detection to predicting the spread of epidemics. Among them, event detection and evaluation in social networks is an attractive and challenging task. A generalized social network refers to a network formed by individuals (people, organizations, or other social entities) connected through a series of social relations (friendship, cooperation, kinship, or information exchange), not limited to popular social media. The occurrence of many events has a huge impact on the corresponding social networks, and the detection and evaluation of these events can provide key materials for many practical applications (such as crisis management and decision - making).
### Limitations of Existing Methods
1. **Methods Based on Text Analysis**: Although widely studied in the field of social media, this method relies heavily on the context provided by users, and there is no available text data for many social interactions (such as friendship, cooperation, and telephone communication).
2. **Methods Based on Graph Feature Statistics**: It is necessary to select appropriate graph features and calculate and compare a large number of statistical variables. Ignoring the dynamic characteristics of network evolution, this method is difficult to evaluate the impact of events in practical applications.
### Solutions Proposed in the Paper
The paper proposes a new hybrid quantum - swarm - intelligence - index method (HQSII), which starts from the perspective of link prediction and includes the Optimal Weight Algorithm (OWA) and the Fluctuation Detection Algorithm (FDA). The main innovations of HQSII are as follows:
1. **Hybrid Index**: For the first time, a hybrid index that can generally describe the evolution of social networks is proposed, exploring the cooperation of different independent similarity indices. The OWA further determines the optimal hybrid index that can achieve higher link - prediction accuracy and better description of network evolution.
2. **Fluctuation Detection**: In order to better avoid the interference of normal network - evolution fluctuations, considering the heterogeneity of micro - node evolution, the FDA for quantifying abnormal fluctuations caused by events is proposed.
3. **Event Detection and Evaluation in Generalized Social Networks**: Based on the OWA and FDA, HQSII can detect events by discovering abnormal fluctuations in generalized social networks and evaluate the impact of events by analyzing the fluctuation trend.
### Experimental Verification
Extensive experiments show that HQSII can accurately detect events and quantitatively evaluate the impact of events in different types of social networks, whether there is a single event or multiple events in the network.
### Conclusion
By proposing HQSII, the paper solves the difficult problems of event detection and evaluation in generalized social networks, especially in the case of large - scale data and limited topological information. HQSII not only improves the accuracy of event detection but also can quantitatively evaluate the impact of events, providing strong support for practical applications.