Natural language or controlled language? The influence of the keyword in the representation for indexing and searching for information

It is an essay to present the debate between natural language or controlled language in the contemporary perspective of the internet and web 2.0 in which the current databases, many of which are open access, such as repositories, are heterogeneous as to the organization and representation of information. The objective is to investigate, through literature review, the use of natural language or language controlled by information systems, in general, and by repositories, in a more specific way, to observe, evaluate and discuss the results, recommendations and reflections. In the qualitative analysis, 90 works were selected and, after previous reading, 73 scientific productions were chosen. Then, the reading was extensive to collect the objectives, methodology, results, conclusions and main points of view of the authors. A representative sample of 38 texts is distributed in 12 categories of analysis to facilitate the macro and micro understanding of this debate and its directions. The results showed that researchers over the years, since the first databases appeared, continue to carry out exploratory studies to observe behaviors of indexers and users, use of tools, processes, products and environment with different contexts and textual typologies. Studies on the application of new methodologies, processes and products are also found and demonstrate new possibilities for reconciling solutions. It was concluded that the coexistence between natural language and controlled language is necessary in information systems for representation in indexing and searching.