Summary
The purpose of this chapter was to cover different topics around time series, word embedding, sentiment analysis, graph theory, and anomaly detection. It's worth noting that the tweets used to illustrate the examples in no way reflect the authors' own opinions: "Whether or not America will be great again is out of scope here":(:( - sarcasm or not?
In the next chapter, we will cover an innovative approach to detect trends out of Time Series data using the TrendCalculus method. This will be used against market data, but can easily be applied in different use cases, including the Sentiment Time Series we built here.