6.3.1 Co-Occurring Terms

In CiteSpace, the word term refers to noun phrases extracted from the text of a bibliographic record or a fulltext document. You can generate a series of networks of co-occurring terms by following the steps below:

Term Type: Noun Phrases

Upon selecting the Noun Phrases radio button under the Term Type panel, CiteSpace will ask you to choose whether to create Part-of-Speech (POS) tags. If tags already exist for your project, it will ask whether you want to use the existing result or re-run the process. Whenever possible, it is a good idea to re-run the process, especially after upgrading CiteSpace to a more recent version.

After choosing the Create POS Tags, CiteSpace will pre-process data files. It will report in the Space Status window how many years of data and how many unique records are valid in your data directory. The figure below shows that 2243 unique records are valid for the dataset Broader Context of CiteSpace, which contains papers that cite the three seed publications of CiteSpace.

Next, select the Term radio button on the Node Types panel. The Detect Bursts button would be enabled automatically; however, since we have not extracted the noun phrases yet, wait until the extraction process complete to use either of the Detect Bursts or the Entropy button.

Now, start the noun phrase extraction process by pressing the GO! button. The extraction process usually takes a while to complete. The status of the process is reported in the Space Status panel, shown in the following figure.

Once the extraction process is completed, the usual dialog box will pop-up for you to choose whether to Visualize the generated network of co-occurring terms, which means terms that appear in the same title, within the same abstract, or in the same keyword list will be connected in the network.

You may apply other analytic functions in CiteSpace to the term network, for example, clustering and labeling.

Now that we have extracted the noun phrases, we can go back to the main interface and select the Entropy function. The result is shown as follows.