Automatic of Indexing: evidence and trends of scientific production indexed in the Brapci

Abstract

Objective: The activities that professional Librarians perform are adapting to new forms of communication and the modern demands of users. Among these activities, there is indexing. This article aims to describe the panorama of scientific production and research trends on studies aimed at automatic, semi-automatic and computer-aided indexing, indexed in the Reference Database for Journal Articles in Information Science (Brapci), in the period from 1972 to 2021. Methods: It is characterized as exploratory and bibliographical with descriptive purposes, since it describes and details the scientific production on the topics "automatic indexing", "semi-automatic indexing" and "machine-aided indexing" indexed in Brapci. Results: Through the analysis of this corpus, it was possible to identify the researchers who produced the most on the subject in question, as well as the journals that most deal with that subject, in addition, the most fruitful periods of time for the production of related works with the automation of indexing and the keywords used to index the analyzed corpus, which, together with the abstracts, signaled the trends of studies in this area. Conclusions: There is a tendency for studies aimed at fully automatic indexing, emphasizing semantic and not just statistical methods, making use of resources derived from Artificial Intelligence. Few proposals for semi-automatic indexing were evidenced. Promising proposals in the practice of thematic representation of information automatically in short spaces of time are presented today.