Linguistic data mining with complex networks: a stylometric-oriented approach

Abstract

By representing a text by a set of words and their co-occurrences, one obtains a word-adjacency network being a reduced representation of a given language sample. In this paper, the possibility of using network representation to extract information about individual language styles of literary texts is studied. By determining selected quantitative characteristics of the networks and applying machine learning algorithms, it is possible to distinguish between texts of different authors. Within the studied set of texts, English and Polish, a properly rescaled weighted clustering coefficients and weighted degrees of only a few nodes in the word-adjacency networks are sufficient to obtain the authorship attribution accuracy over 90%. A correspondence between the text authorship and the word-adjacency network structure can therefore be found. The network representation allows to distinguish individual language styles by comparing the way the authors use particular words and punctuation marks. The presented approach can be viewed as a generalization of the authorship attribution methods based on simple lexical features. Additionally, other network parameters are studied, both local and global ones, for both the unweighted and weighted networks. Their potential to capture the writing style diversity is discussed; some differences between languages are observed.

References in corpus (16)

• Fast unfolding of communities in large networks
• The structure and function of complex networks
• The architecture of complex weighted networks
• A survey of cross-validation procedures for model selection
• Characterization of complex networks: A survey of measurements
• Generalizations of the clustering coefficient to weighted complex networks
• Statistical analysis of 22 public transport networks in Poland
• Weighted Assortative And Disassortative Networks Model
• Text authorship identified using the dynamics of word co-occurrence networks
• Generalized Hurst exponent and multifractal function of original and translated texts mapped into frequency and length time series
• In narrative texts punctuation marks obey the same statistics as words
• Modeling the average shortest path length in growth of word-adjacency networks
• Complex network analysis of literary and scientific texts
• Microscopic activity patterns in the Naming Game
• Evolution of Vocabulary on Scale-free and Random Networks
• Hierarchical organization of H. Eugene Stanley scientific collaboration community in weighted network representation

Cited by in corpus (12)

• Using word embeddings to improve the discriminability of co-occurrence text networks
• Complex systems approach to natural language
• Universal versus system-specific features of punctuation usage patterns in~major Western~languages
• Text characterization based on recurrence networks
• Statistics of punctuation in experimental literature -- the remarkable case of "Finnegans Wake" by James Joyce
• Multifractal hopscotch in "Hopscotch" by Julio Cortazar
• DHLP 1&2: Giraph based distributed label propagation algorithms on heterogeneous drug-related networks
• Punctuation patterns in "Finnegans Wake" by James Joyce are largely translation-invariant
• Link Prediction via controlling the leading eigenvector
• Language Networks: a Practical Approach
• Average shortest-path length in word-adjacency networks: Chinese versus English
• God ($\equiv Elohim$), the first small world network