Characterization of literary works using co-occurrence networks

• Grant number: 17/26981-9
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: April 01, 2018
• End date: March 31, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Jesús Pascual Mena Chalco
• Grantee: Bruna Pereira Santos
• Host Institution: Centro de Matemática, Computação e Cognição (CMCC). Universidade Federal do ABC (UFABC). Ministério da Educação (Brasil). Santo André , SP, Brazil

Abstract

The concept of networks can be employed to different daily events, such as telecommunications, transport, social connections and even narratives that constitute literary works (compositions or documents). Amongst different types of existing words network, there is a type called co-occurrence network, relating the vertices (words) from their proximity in a text. In this scientific initiation project, the application of the co-occurrence network concept in literary works will be investigated for one year. In this context, the following steps will be performed: (I) removal of blanks, punctuation marks, as well as the stop words; (II) automatic text normalization; (III) construction of co-occurrence networks, and (IV) comparison and characterization of networks representing literary documents. Since there is currently a challenge in the classification of several types of texts, the importance of this project lies in showing that the application of graph theory can contribute to several quantitative analyzes of textual documents, supporting their characterization. Some of the possible analyzes are the discovery of the thoughts (concepts) most surrounded and the exposure of the power law phenomenon in words distribution, as well as the network characterization through structural properties. (AU)