Differences between normal and shuffled texts: Structural properties of weighted networks

Advances in Complex Systems (Impact Factor: 0.97). 11/2011; 12(01). DOI: 10.1142/S0219525909002039


In this paper we deal with the structural properties of weighted networks. Starting from an empirical analysis of a linguistic network, we analyze the differences between the statistical properties of a real and a shuffled network. We show that the scale-free degree distribution and the scale-free weight distribution are induced by the scale-free strength distribution, that is Zipf's law. We test the result on a scientific collaboration network, that is a social network, and we define a measure – the vertex selectivity – that can distinguish a real network from a shuffled network. We prove, via an ad hoc stochastic growing network with second order correlations, that this measure can effectively capture the correlations within the topology of the network.

Download full-text


Available from: Geoff Rodgers
  • Source
    • "when sentence structures are randomized, predicts a distribution of degrees that mirrors Zipf's law [8] [10]. That degree distribution includes hubs reminiscent of function words [11]. "

    Full-text · Article · Aug 2014 · Physics of Life Reviews
  • Source
    • "It is observed that the networks from randomized texts also exhibit small-world and scale-free characteristics. Masucci and Rodgers showed [8] [9] that the power law distribution holds when they shuffled the words in the text. Thus, they showed that degree distribution is not the best measure of the self-organizing nature of weighted linguistic networks. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper studies the properties of the Croatian texts via complex networks. We present network properties of normal and shuffled Croatian texts for different shuffling principles: on the sentence level and on the text level. In both experiments we preserved the vocabulary size, word and sentence frequency distributions. Additionally, in the first shuffling approach we preserved the sentence structure of the text and the number of words per sentence. Obtained results showed that degree rank distributions exhibit no substantial deviation in shuffled networks, and strength rank distributions are preserved due to the same word frequencies. Therefore, standard approach to study the structure of linguistic co-occurrence networks showed no clear difference among the topologies of normal and shuffled texts. Finally, we showed that the in- and out- selectivity values from shuffled texts are constantly below selectivity values calculated from normal texts. Our results corroborate that the node selectivity measure can capture structural differences between original and shuffled Croatian texts.
    Full-text · Article · May 2014
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we analyse the street network of London both in its primary and dual representation. To understand its properties, we consider three idealised models based on a grid, a static random planar graph and a growing random planar graph. Comparing the models and the street network, we find that the streets of London form a self-organising system whose growth is characterised by a strict interaction between the metrical and informational space. In particular, a principle of least effort appears to create a balance between the physical and the mental effort required to navigate the city.
    Full-text · Article · Mar 2009 · Physics of Condensed Matter
Show more