Domain analysis of scientific production in Information and Communication Technology in the context of small farmers

Abstract

1 Introduction This research presents an evaluation of the scientific production which addresses the issue involving the use of Information and Communication Technologies (ICT) in the context of small rural properties. These studies will be conducted by locating the core of researchers and ideological currents. Therefore, it aims to define the structure in which the researchers that deal with this subject are organized using domain analysis- HJORLAND (2002); MAI (2005); MIGUEL, MOYA-ANEGON and HERRERO-SOLANA (2008) carried out using the application of the analysis of citation and cocitation of authors - McCAIN, (1990); GLANZEL (2003); OLIVEIRA and GRACIO (2013), identifying the main representatives in the area (citation) and the interlocution between them (cocitation). 2 Research material and method The identification of the articles that composed the analysis was done in the SCOPUS database in September / 2017, using the terms [small farmer AND ICT]. The articles were selected by their summary in order to confirm their relevance to the topic (Table 1 - APPENDIX A). Once the article was considered appropriate, the set of citations of each one was downloaded through files in the Comma-Separated Values (CSV) format. The set of citations collected were consolidated into a text file and the analysis process was initiated involving the exclusion of all information from the reference, keeping only the authors. This process also involved an analysis of references that contained multiple authors removing double citations. A count was conducted to identify the most common citations. In total, the research consisted of 471 researchers referenced in 15 articles.The researchers selected were those who obtained more than one citation to compose the study of cocitation. In order to underlie the results with a qualitative approach, information about these researchers was sought in academic and professional social networks (Research Gate, LinkedIn), educational institutions and Wikipedia. In order to develop the cocitation network, an asymmetric matrix of the citing articles was first constructed with the most cited researchers, annotating their citations (denoted by number 1) or not (denoted by the number 0) of the citing work for the cited researcher. From this matrix the symmetric matrix was generated with the absolute frequencies of cocitation among the most cited authors. The network of cocitation was elaborated from the symmetric matrix using the statistical analysis software UCINET and its visualization tool NetDraw, demonstrating the most similar authors related to the absolute values of cocitation. 3 Results A temporal cut was not made for the selection of the corpus analyzed. Instead, this research chose to consider the search only for the subject researched. Nevertheless, there were a limited number of articles (15), all of them published since 2002. The authors were selected based on the references in these articles. These authors were those who were mentioned in at least 2 of them, which resulted in a total of 24 authors. It is noteworthy that two-thirds of the articles analyzed cited at least one of these 24 researchers, proving their knowledge in the subject analysed. Table 2 - APPENDIX B presents the most cited authors, their respective nationalities and institutional associations. The three most cited authors - with three mentions each - are: Richard Heeks (England), from the University of Manchester, which carried out research about the use of ICT for international social development - ICT4D; Keith Griffin (Panama), Magdalen College, worked on the economic aspects of poverty reduction; and Ashok Jhunjhunwala (India), of Department of Electrical Engineering of the Indian Institute of Technology, articulated projects involving technology-based companies. Figure 1 - APPENDIX C presents the citation network with the most cited authors, in which the size of the squares reflects the intensity of their level in the network. The thickness of the lines is proportional to the frequency of cocitation between the authors they connect. In the complete article the analysis of this cocitation network is deepened using the description of the identified clusters and with the calculation of social network indicators (OTTE; RSOUSSEAU, 2002), as the indicator of Network Density and Centrality of Level. 4 Final ConsiderationsSome considerations arose from the results obtained mainly related to the identification of the three clusters that can represent invisible colleges within the analyzed domain: one of them is composed by only two authors from Trinidad and Tobago who are coauthors in a work edited by a third author also among the most cited; another one is formed by the approximation of Indian authors and also Africans and Americans; the third one has a massive presence of European authors who also have associations with Indian authors. A recent growth in the interest on the subject researched is noticed, despite its little consolidation. It’s represented by the increase in publications since the 2000s.