Create Text Network in a Few Steps

No-Code Text Analysis

In text analysis, researchers often focus on which keywords appear most frequently. However, meaning in academic text is rarely carried by individual terms alone. It is largely shaped by how words co-occur within shared contexts.

A co-occurrence network captures this relational view of text. It represents text as a network in which nodes correspond to words (or keywords), and edges represent their co-occurrence within a defined context such as a sentence, document, or sliding window. This approach allows researchers to analyze the structural relationships among concepts rather than treating terms as independent units.

In this tutorial, we show how to construct a co-occurrence network from academic text data using terminology commonly adopted in the research literature.

This series introduces how to collect and analyze unstructured text data in NetMiner 5. ☞ Start NetMiner 5 Trial

What will you learn in this tutorial?

• How to build keyword co-occurrence relationships from text data.

Who is this for?

• Researchers interested in analyzing text as a network of interconnected keywords.
• Those studying discourse, topics, or conceptual structures in text.
• Humanities and social science researchers who want to move beyond frequency analysis to structural interpretation.

Which NetMiner 5 feature is used?

• Word co-occurrence network (Preprocess > Text > Word Network > Sliding Window)
• Centrality analysis (Network > Centrality)

In this tutorial, we use a sample dataset of AI academic publications. (File Location: \SampleData\Text Analysis\ENG\Bibliography) ☞ See Description of NetMiner sample data

1. Building a Keyword Co-occurrence Network

Go to Preprocess > Text > Word Network > Sliding Window.

NetMiner provides two approaches for constructing a keyword co-occurrence network.

• Sliding Window Edges are created only when words appear close to each other in the text. Co-occurrence is defined by distance, measured as the number of words between terms. This distance is controlled by the window size.
• Co-occurrence All words that appear within the same unit (document, paragraph, or sentence) are connected. Word distance is ignored as long as the terms occur in the same text unit.

Next, select the Input data on the right panel.

• If your input is raw text, select Raw Text.
• If your input is a bag-of-words representation, select TDM (Term-Document Matrix). (*If you select TDM, you can skip steps 1.1–1.2.)

In this example, we select Raw Text.

Under Data Item, choose the imported dataset. Under Attribute, select the column that contains unstructured text.

1.1 Morphological Analysis

Next, set the language for morphological analysis, choose the parts of speech to extract, and define the processing unit. For English, check Case Sensitive if you want to distinguish uppercase and lowercase words.

1.2 (Optional) User Dictionary

If you would like to apply a user dictionary, click the icon on the right under User Dictionary. ☞ Learn more about User Dictionaries

1.3 Word Co-occurrence Network

Finally, configure the settings for constructing the co-occurrence network.

• If you set the Window Size to 2, edges are created only between immediately adjacent words in the text.
• The Link Frequency Threshold determines the minimum number of co-occurrences required for an edge to be included in the network. For example, if you set the threshold to 2, only word pairs that co-occur at least twice will be retained.

Click Run Process to generate the network.

2. Results

Click the [T] Word Network tab to inspect the word association data.

At the bottom right, click Add to Dataset to save the resulting keyword co-occurrence network to your dataset.

3. Identifying Key Terms in the Semantic Network

To find centrally positioned words in the network, go to Network > Centrality > Betweenness Centrality.

Betweenness centrality measures how often a node lies on the shortest paths between other nodes. Keywords with high betweenness centrality often connect different topics or mediate the overall flow of meaning in the text.

Select the target network (here, the keyword co-occurrence network added in Step 2) and click Run Process.

The results report shows the top keywords by centrality, along with the Centralization Index, which indicates how strongly importance is concentrated among a few keywords.

Identifying highly central keywords in a keyword co-occurrence network helps reveal the core flow of meaning and the underlying structure of a text, extending interpretation beyond simple frequency-based analysis.

Spend less time analyzing data. Spend more time on insight.

**☞ Try for Free ☞ NetMiner Help Center**

NetMiner | CYRAM Network Science · Social Network Analysis · Graph Analytics https://www.netminer.com