River Sentiment Dashboard

People talk about rivers online: from complaints about pollution to celebrations of wildlife, rivers provoke passionate social media comments. When people express their feelings about rivers on Twitter, we get a glimpse of how nature affects human wellbeing. The River Sentiment Dashboard displays social media sentiment alongside data about the ecological status of more than 450 rivers in the Thames basin in England. This prototype has been developed by Oxford University and Thames21.

How do you feel about your local river?

Find on a Map

How people feel about...

This graph shows how positively or negatively people felt about a river over time. When the line goes up, people expressed positive feelings about the river. If it goes down, then they expressed negative feelings. Using natural language processing on millions of tweets, we assign each tweet mentioning a river by name a sentiment score between 5 (very positive) and minus-5 (very negative). The graph displays the monthly mean score. Click and zoom to increase the temporal resolution to daily view. Double-click to return to the default view. Hint: when you tweet about a river, make sure to mention its full name. Then our algorithm is more likely to pick up your tweet.

View Raw Data

Basic emotions associated with...

This bar chart shows the kinds of emotions people express when tweeting about a river. For instance, the words people use to tweet about pollution might be associated with disgust or the words they use about wild swimming might be associated with joy. The data show the percentages of eight basic emotions detected in any tweets mentioning a river by name. We use the NRC Word-Emotion Association Lexicon to detect anger, fear, anticipation, trust, surprise, sadness, joy, and disgust in English-language tweets.

View Raw Data

Common phrases associated with...

This word cloud shows the most common phrases people use when tweeting about this river. The results give an indication how frequent certain topics come up in connection with a river. We conduct frequency analysis to count the distribution of noun phrases (n-grams) in all tweets mentioning a river by name.

View Raw Data

Ecological status of...

This table shows how the Environment Agency assessed the environmental quality of the river over time. Ecological status can be either high, good, moderate, poor or bad. According to the EA their assessment of ecological status is based on biological quality, general chemical and physico-chemical quality, water quality with respect to specific pollutants, and hydromorphological quality. The data shown are from the Catchment Data Explorer.

Why is ...not achieving good status?

This diagram shows what keeps the river from achieving good ecological status. The diagram shows the relative responsibility of certain sectors of society (left) for the environmental issues (right) that stand in the way of improving the quality of the river. We plot Environment Agency data of the “reasons for not achieving good (RNAG)” from the Catchment Data Explorer.


Sentiment Polarity Score

Negative (-1.1 to 0.26)
Neutral (0.26 to 0.4)
Positive (0.4 to 2.9)

The Sentiment Polarity Score measures how positively or negatively people talk about rivers on Twitter. The score can have a value between -5 (extremely negative) and 5 (extremely positive) but will usually fall somewhere in between. Negative sentiment is coloured red, neutral sentiment is coloured yellow and positive sentiment is coloured blue. The score is based on the average of all tweets mentioning the river over 10 years.