topics review.What can we do with Topic Mining?We have seen how you can implement topic modelling, but often the question is: Why should I do this in a business?Applying topic mining to your customer or internal data can easily and quickly enable your analytics teams (even the domain experts and business analysts) to visually identify causalities and correlations in the data, and help them solve challenging business problems like customer churn and employee satisfaction.The most popular of application are:Social Media Review AnalysisCase System HandlingNPS Interview AnalysisLet’s have a look at some of the most common ways to move forward from here.Drill DownHaving topic assigments at hand you can put this into common BI-Tools..Is there a specific region where topics cluster?.Did we have a problem on a specific weekend?.All of these questions can be answered retrospectivly from the data.Analyze the Trend of TopicsSometimes you are interested in single topics..You want to know if the “delivery time” topic is raising over time..By doing a time series forecast on the number of documents per day associated to a topic, you can create an alarm signal..The signal notifies the user or responsible business unit that there is a spike in “delivery time” as topic and possible bottle-neck in the delivery process, allowing the business to resolve this issue and maintain good service.Preprocessing for further AnalyticsOf course, you can use these topic distributions now to further analyze your text data:Use the distribution of topics for a product as an input for further analysis — e.g..supervised learning.Topic distribution for clustering to find similar products.Anomaly detection on the topic distribution and on further attributes to find “odd” products.ConclusionIn this article we’ve used Topic Mining on Amazon Reviews..As you can see this technique can be used very quickly and generates understandable results..All the processes, scripts and the sample of the data is available on my github profile.. More details