It is important to evaluate the results of your actions to determine the validity and quality of your analysis and to plan the next cycle of analysis.
After you take actions based on the insights that you discovered, verify that the effect of your actions correspond to what you expected to occur. For example, assume you analyzed car complaint data. By using the content analytics miner, you identified a rapid increase in the number of complaints about a specific car model with a specific problem. Based on your analysis, you identified different facets that are strongly correlated to this problem and such facets indicated certain solutions and actions to take.
If the effect of your actions is different than what you expected, you can investigate if there are any other external factors that might be related to the increase in complaints. For example, the changes and deviations in the data that you observed might be due to a change in the data input operations or a database problem that caused duplication of the data. Alternatively, there might be a rumor about a potential problem and many people called to ask about the rumored problem. Validation of the effect of your actions can lead to improved analysis and the setting of new objectives for the next cycle of data analysis.
You can use the content analytics miner to validate the effect of your actions. If you save the query that helped you discover the problem, you can run the query periodically after new data is added to the collection. For example, run the query and view results in the Time Series view. If your actions are effective, then the frequency counts should decrease over time.
