Tutorial 9 - Discourse Analytics

Word Frequency

This report analyzes a group of student’s online discussion around the topic of Climate change. The report includes analysis of word frequency using token, bi-gram, tri-gram and tfif (Term Frequency - Inverse Document Frequency), and also the overall sentiment of discussion sentences. The data has been processed by removing stop words, as stop words are used more frequently than other words, but does not provide as much value as context specific terms.

Loading...

The top 4 frequent word in the discussion is “climate”,“change”, “course” and “assignment”, which aligns with the the context. Suggesting students focuses discussion around the main topic, whilst also discusses a lot surrounding the course content.

Bi-Gram

Loading...

The bi-gram results aligns with the single word frequency chart above, whereby “climate change” being the top phrase suggests the students are mostly on topic in discussions. Other related phrases such as “global warming”, “sea level” and “level rise” also reinforces this observation. While other phrases (“thank much”,“hope helps”) also reinforces the polite environment ion the MOOC.

Tri-Gram

Loading...

From the tri-gram we can see that the narrative remains the same, whereby the phrases are related to: 1.Climate change, 2. Helpful comments 3. Course content.

Loading...

This chart above highlights the tokens with the highest TF-IDF scores(words that are less common but particularly informative). Terms like “extent”, “shrinking”, “hiatus”, and “glaciers” suggest a more technical or scientific focus within certain documents. These words continue to reflect the focused discussion around climate-related, showing the specialized and more analytical contributions students engages in.

Sentiment Analysis

Loading...

This sentiment analysis chart shows the distribution of emotional tone across individual entries. Most entries fall into the positive or mildly positive range, with a few displaying neutral or slightly negative sentiment. This trend suggests that the overall tone of the dataset is constructive and supportive, which aligns well with educational or academic discourse settings where respectful and informative exchanges are encouraged.

Export Chart Image
Output Format
PNG SVG
Background
Set background transparent
Size
Width (Pixel)
Height (Pixel)
Pixel Ratio