Bibliometric analysis

PDFs to a word
co-occurrence network in VOSviewer

Upload your PDFs, extract the text, generate a JSON word co-occurrence network file and open it directly in VOSviewer Online.

VOSviewer JSON Word co-occurrence Stopwords PT + EN 100% local

1 — Upload PDFs

Drag PDFs here

or click to select — multiple files supported

2 — Network parameters

Minimum frequency
3

Minimum number of occurrences of a word

Minimum co-occurrence
2

Minimum link strength between two terms

Maximum words
100

Most frequent terms to include in the network

Minimum length
3

Minimum number of characters per word

Words to exclude (custom stopwords)

Type and press Enter, or paste separated by comma, semicolon or space

or paste a block

How to use

01

Upload the PDFs

Drag or select one or more PDFs. The text is extracted locally, with nothing sent to a server.

02

Adjust the parameters

Set the minimum frequency, minimum co-occurrence and maximum number of terms to control the size of the network.

03

Open in VOSviewer

Download the generated JSON and drag it into VOSviewer Online. The co-occurrence map will be displayed immediately.

Frequently asked questions

What is a co-occurrence network?

It is a network where each node represents a word and the links indicate that two words appear together in the same sentence. The more often they co-occur, the stronger the link. This is a classic analysis in bibliometrics and text mining.

How do I use the JSON in VOSviewer Online?

Download the generated JSON file, go to app.vosviewer.com, click "Open" in the top right, select "VOSviewer JSON file" and load the file.

Do scanned PDFs work?

Not directly. PDFs made up of images only need OCR first. The tool extracts text only from PDFs with a text layer (born-digital or with OCR already applied).

Which stopwords are removed?

Function words in Portuguese (de, a, que, para, com, etc.) and English (the, and, for, of, in, etc.) are removed automatically. This ensures that only terms with semantic content appear in the network.

The network produced no links — what now?

Try reducing the minimum frequency and minimum co-occurrence parameters. For short texts, values of 1 or 2 are usually enough. Increasing the maximum number of words also helps.