Aspect-based Sentiment Analysis
Determine the sentiment attached to entities within the text.
Aspect-based Sentiment Analysis, also referred to as Targeted Sentiment Analysis and Opinion mining, identifies entities within the text and measures the sentiment surrounding them. This demo utilizies AWS's Comprehend and Azure's Cognitive Services. We find this AI technology doesn't currrently work well on more complex pieces of text with more nuanced language.
- Enter some text in the box below, or select from one of the example texts.
- Click the Analyse button.
- Select the Azure, and AWS tabs to view the aspect-based sentiment analysis results.
More information about this demo
This service is inlcudes both AWS Comprehend and Azure's Cognitive Services. Whilst identifying entities within the text, these tools also measure the sentiment surrounding each entity. The sentiment can either be "positive", "negative" or "neutral".
Things to consider
Sentiment Analysis has a wide range of potential uses. An organisation that values public sentiment or attitude towards them could use sentiment analysis tools to analyse social media, blog posts or discussions on online forums. Or universities could use sentiment analysis to analyse student feedback and comments, either from their own surveys or from online sources such as social media. Targeted Sentiment analysis allows you to gain further detail and insight into feedback, highlighting specific items with sentiment attached to them.
It is still worth noting that AI based text analysis tools have difficulty recognising more subtle and complex aspects of language like sarcasm, irony, negations and jokes. We have found this particularly significant when looking at aspect-based sentiment analysis. When looking at the data returned from the AI tools included within this demo it is quite difficult to gain useful insight on data beyond very basic text. Significant human input would be required to decifer whether the returned analysis was accurate and insightful.
This technology is relatively new and we expect it's accuracy and ability to analyse more complex text to improve over time.