学术文献库

In this paper we extrapolate the information about Bible's characters and places, and their interrelationships, by using text mining network-based approach. We study the narrative structure of the WEB version of 5 books: the Gospel of Matthew, Mark, Luke, John and Acts of the Apostles. The main focus is the protagonists' names interrelationships in an analytical way, namely using various network-based methods and descriptors. This corpus is processed for creating a network: we download the names of people and places from Wikipedia's list of biblical names, then we look for their co-occurrences in each verse and, at the end of this process, we get N co-occurred names. The strength of the link between two names is defined as the sum of the times that these occur together in all the verses, in this way we obtain 5 adjacency matrices (one per book) of N by N couples of names. After this pre-processing phase, for each of the 5 analysed books we calculate the main network centrality measures (classical degree, weighted degree, betweenness and closeness), the network vulnerability and we run the Community Detection algorithm to highlight the role of Messiah inside the overall networks and his groups (communities). We have found that the proposed approach is suitable for highlighting the structures of the names co-occurrences. The found frameworks' structures are useful for interpreting the characters' plots under a structural point of view.