This chapter introduced question answering applications, one of the most popular uses of the Web. The popularity of the Stack Exchange network, and of the Stack Overflow website for programmers, is driven by the high-quality material curated by the community. In this chapter, we discussed how to interact with the Stack Exchange API and how to access the whole data set from Stack Exchange using their data dumps.

The second part of the chapter introduced the task of classification and the related supervised machine learning approaches to tackle it. The availability of labeled data from Stack Exchange has provided the opportunity to build a predictive model. The use case proposed in this chapter was the prediction of labels to be used as question tags, but the techniques can be applied to a variety of applications. The final part of the chapter extended the discussion, showing how a machine learning model can easily be integrated in a real-time user-facing application.

The next chapter...