It is obvious that there is an increasing interest in network research since the early 1970s. The network perspective has become more popular in a number of social science disciplines because it enables researchers to model the relationships between actors. The aim of this study is to investigate the evolutionary development of social network theory and social network analysis. To detect the path of evolution we used papers published in the Social Networks Journal from 2000 to 2019 inclusive which is one of the pioneer journals of this field. Up to emerging big data the only way to understand the text was reading by eye. However, there is another alternative. Text mining is a bunch of approaches and instruments for analyzing the enormous amount of textual data. These instruments provide benefits like summarizing, classifying, visualizing the text. Text mining leans on breaking the sentences into words and calculating their occurrence for each document. This process helps to build an enormous matrix for seeking inter-conceptual relations. In this study, we employed text visualization to extract information from 766 articles from the Social Networks Journal. We used R language and its powerful visualization and reporting tool Shiny. We developed an open, interactive and flexible interface to investigate what was the real topic in the papers. The name of this interface is JournalAnalytics. By this interface, we could extract the most frequent words and bigrams (the phrases) according to the year. We also used world co-occurrence network analysis which combines word co-occurrence analysis with network analysis based on graph theory. It was possible to constitute word co-occurrence network hence we detected major groups of words that represent the intellectual structure of social network theory and analysis. In addition, Sankey diagrams were used to show the power of relations between co-occurrent words.