Sentiment Analysis

During our discussions on the way the mood tree works, we realised that we did not want to limit the data collection. If the tree only captured the moods of the individuals who visited it, then it would not be representative of the neighbourhood or city. There wouldn't be a lot of colours displayed in that case. Therefore, we thought we could leverage sentiment analysis in addition to the physical participation at the tree. Sara showed us a website that performs sentiment analysis by crawling through social media threads and public feeds to capture the feelings of people about particular topics.

Opinion Crawl: http://www.opinioncrawl.com/

The results after searching brexit

A couple of weeks after this discussion, I attended an event by Amazon Web Services on customer experience and innovation. During the event, we had the opportunity to discuss the latest technologies available (VR, AR, ML, Lex,..) and also brainstorm on how these can be implemented in various industries and companies. In the context of call centres, someone from Amazon Connect mentioned the new technologies their clients now use called “transcribe” and “comprehend”. Transcribe enables conversations over phone to be written out. Comprehend feature on the other hand scans transcripts, emails coming in, social media threads and many more and analyses what people are talking about. It can then identify the most common topics discussed by people and perform sentiment analysis.

Here is a video that describes the “comprehend” feature:

Hearing how Amazon is using natural language processing to perform sentiment analysis supported our initial discussions on how we could leverage this in our design solution. We thought the tree could perform natural language processing and sentiment analysis of individuals that visit the tree. In the absence of visitors, it could capture moods and feelings from the public threads.

All of these discussions made us think about "language" and "sound" at a holistic level. We appreciated the opportunities that language carries and how it could be leveraged in different ways.