Font Size: a A A

Study On Information Extraction Based On Complex Network Invulnerability

Posted on:2017-02-10Degree:MasterType:Thesis
Country:ChinaCandidate:S Q LiFull Text:PDF
GTID:2180330485989968Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Invulnerability of complex networks is an important content in the study of complex networks. In different areas and systems, detailed definition of invulnerability of complex network has a certain difference. In a large amount of literature, most of it is the research on the algorithm of the measurement of the invulnerability, but in specific areas on the basis of invulnerability, less of important information analysis is extracted. This paper is mainly based on the characteristics and invulnerability measurement of complex networks, study on the extraction of important information in the Chinese word co-occurrence network and urban public transport network.The architecture of this paper includes: First, introduce complex network and its invulnerability from the theory of comprehensive and systematic, then select two typical areas: Chinese word co-occurrence network and urban public transport network to carry on the correlation analysis, get the key information from the data. Such as the following:(1)Analysis of the complexity and invulnerability based on the Chinese word co-occurrence network. First, the four Chinese word co-occurrence networks are constructed based on the corpora of the 《People’s Daily》 made by Peking University Institute of Languages and modern Chinese of The National Language Committee. The indices are calculated and the distribution curves are plotted by the results. It is proved that the word co-occurrence networks followed with small-world effect and scale-free property. For the above word co-occurrence networks, carry out the analysis of invulnerability. The indices to measure the invulnerability of complex network are the decline rate of maximum connected sub-graph size and overall efficiency. The experiment results show that the entire network rupture into several isolated network with the node of great degree and betweenness in the word co-occurrence networks removed. The article will become fragmented. It reflects laterally the core vocabulary in the Chinese words. Next, it provides a solution to study on the core vocabulary in Chinese words.(2)Analysis of the complexity and invulnerability based on urban public transport network. This paper selects the data of Beijing city transit network on the official website, first, analyze the data source, build a network of bus lines, use the MATLAB software to sum up the basic statistics, get the scale-free characteristic of Beijing city public transport network according to the cumulative degree distribution curve and the double logarithmic accumulation degree distribution. Verify the distinctive small world of the network, according to the comparison of the average shortest path and clustering coefficient. Finally, select the deliberate and random two attack models to analyze invulnerability of the urban public transport network. According to the results of the above analysis, we can quickly and directly extract the level of importance of each bus station in urban public transport network, then we can focus on the security of key sites.
Keywords/Search Tags:complex network, invulnerability, Chinese word co-occurrence, urban public transport network
PDF Full Text Request
Related items