**3. Methodology and recovery documents**

### **3.1. Database**

The research was based on the selection of scientific publications collected in ISI/Web of KnowledgeSM database. It is international bibliographic data and is used as reference for the generation of indicators of S&T and Innovation [2].

### **3.2. Strategies for recovery documents**

The strategy for the recovery of publications on a particular scientific topic must be carried out carefully to avoid damaging in the analysis of scientific literature. For the analysis of results in an accurate overview of the subject, ideally, the "search expression" must promote the recovery of all relevant publications on the subject present in the database, and at the same time, exclude nonrelevant publications. However, such situation is difficult to achieve since the multidisciplinary issues receive contributions from several areas of knowledge.

The solution for recovery of significant and coherent set of publications that allows an analysis of the scientific production in the biotechnology area was the elaboration of a complex search expression, consisting a large of set selected and tested keywords.

At first, we tried to develop search strategies in the mode "ISI-general search," using keywords from the available literature and descriptors present on the *Platform-Lattes*/National Council for Scientific and Technological Development (CNPq) - Brazil and *Portal Innovation*/ MSCT-Brazil related to research groups and ongoing research in the biotechnology area. These search expressions were just an exploratory character, that is, serve as a basis for elaborate more refined expressions and evaluate how the area is organized into database [20]. A selection from bibliographic references was also used.

In order to achieve the purpose of mining scientific production on biotechnology in Brazil over a period of 21 years (1995–2015), 60 distinct descriptors (**Table 1**) were used, selected by experts in biotechnology area, interviewed and research groups in the biotechnology area sought in the base *Plataforma Lattes CNPq/Brazil* and *Portal Innovation*/MSCT-Brazil.

From our previous experiences, we found that the best way to do this recovery would be a combination of different expressions. For this, we used the advanced search mode of ISI/Web of Science and the advanced search. Once defined the strategy, we used all these keywords, and various search expression and their combinations in the database ISI/Web of Science only looking for indexed publication using document types "article." The fields "Title, Summary, and Keyword" were used to analyze the themes of scientific publications with international dissemination [21].


**Table 1.** Biotechnology-related terms used to search articles in database ISI/Web of Science.

were: plant and animal breeding, food technology, bioinformatics and immunology, diagnosis and prospecting on biotechnology, and detection of genetically modified organisms.

The research was based on the selection of scientific publications collected in ISI/Web of KnowledgeSM database. It is international bibliographic data and is used as reference for the

The strategy for the recovery of publications on a particular scientific topic must be carried out carefully to avoid damaging in the analysis of scientific literature. For the analysis of results in an accurate overview of the subject, ideally, the "search expression" must promote the recovery of all relevant publications on the subject present in the database, and at the same time, exclude nonrelevant publications. However, such situation is difficult to achieve since the multidisciplinary issues receive contributions from several areas of

The solution for recovery of significant and coherent set of publications that allows an analysis of the scientific production in the biotechnology area was the elaboration of a complex

At first, we tried to develop search strategies in the mode "ISI-general search," using keywords from the available literature and descriptors present on the *Platform-Lattes*/National Council for Scientific and Technological Development (CNPq) - Brazil and *Portal Innovation*/ MSCT-Brazil related to research groups and ongoing research in the biotechnology area. These search expressions were just an exploratory character, that is, serve as a basis for elaborate more refined expressions and evaluate how the area is organized into database [20]. A

In order to achieve the purpose of mining scientific production on biotechnology in Brazil over a period of 21 years (1995–2015), 60 distinct descriptors (**Table 1**) were used, selected by experts in biotechnology area, interviewed and research groups in the biotechnology area

From our previous experiences, we found that the best way to do this recovery would be a combination of different expressions. For this, we used the advanced search mode of ISI/Web of Science and the advanced search. Once defined the strategy, we used all these keywords, and various search expression and their combinations in the database ISI/Web of Science only looking for indexed publication using document types "article." The fields "Title, Summary, and Keyword" were used to analyze the themes of scientific publications with international

sought in the base *Plataforma Lattes CNPq/Brazil* and *Portal Innovation*/MSCT-Brazil.

search expression, consisting a large of set selected and tested keywords.

selection from bibliographic references was also used.

**3. Methodology and recovery documents**

generation of indicators of S&T and Innovation [2].

**3.2. Strategies for recovery documents**

192 Knowledge Management Strategies and Applications

**3.1. Database**

knowledge.

dissemination [21].

The software VantagePoint® was used to perform the processing of information from articles retrieved in order to expand and enrich the results. VantagePoint® allows management of big data and information in order to present correlations of distinct variables of interest. It also allows us to identify Who, What, When, and Where, help us clarify relationships and find critical patterns, among other possibilities [2].

For visualization of the data, we used the software VOSviewer®. It is a latest free software for the representation and analysis of information, which appears as an alternative to the traditional techniques of multidimensional representation and network display. VOSviewer combines the visualization and clustering techniques, favoring analysis while bypassing unnecessary technical complications. It make possible to view the maps of collaboration between institutions, countries, and keywords [22].

After this step, macroindicators were generated for providing a global overview of the scientific production on biotechnology in Brazil and raising the main issues as follows: (1) the total number of articles published per year (1995/2015) and the trend of publication of the most frequently used terms; (2) major journals that institutions often publish their articles in this specific area; (3) key areas of knowledge and the number of articles published on the used terms; and (4) the number of papers published by institutions and maps of network (institutions and countries) for the purpose of identifying partnership [2].

The trends for biotechnology were mapped out through a set of variables such as the identification of the actors involved as well as the institutional partnerships and networking.
