Establishing Continuous App Improvement by Considering Heterogenous Data Sources

Abstract

<p class="0abstract"><span lang="EN-US">Mobile apps have penetrated the market and are being used every­where. Companies developing apps face increasing challenges such as short time to market or demand for high quality. Furthermore, the success of an app also depends on how users perceive its quality. Feedback provided by users influences other potential users and provides new opportunities for identifying features. Consequently, it is a valuable source of input for app developers with respect to product improvements. One form is textual feedback. This kind of feedback is usually distributed across various data sources. Therefore, it must be captured from these sources and put into one single pool of data before it can be analyzed. The analysis must take into account the peculiarities of the short release cycles and high change rate of features for mobile apps. In this article, we present User Echo Service (UES), which was built to address heterogeneous data sources. The aim of UES is to allow product managers to be able to be always up to date with the latest feedback data. Therefore, we have created an extensible architecture aimed at supporting different data sources and present the feedback collection scheduling system. This forms the prerequisite for subsequent analyses of the collected data. We discuss our solution and provide ideas for future development.</span></p>

The way it made me feel – Creating and evaluating an in-app feedback tool for mobile apps

“…Stade et al [13] presented the FAM E framework for collecting feedback and monitoring data to support requirements elicitation. Scherr et al [9] introduced a common data model for textual feedback from multiple data sources and an architecture for feedback collection. In this work, they considered several feedback data sources, e.g., Apple App Store, Google Play, Facebook, and Twitter.…”

“…For the prototype, we first developed an in-app feedback data model. As we intend to use the tool in UES, the model is based on the feedback data model defined for UES as presented in [9]. According to our model, in-app feedback can include multiple ratings, textual feedback, and attachments.…”

“…This is why we want to apply the crowd idea to quality assurance as well and help developers understand feedback trends in their products without spending much effort. This means user feedback should be understood as a continuous stream of data, as users are constantly changing their feedback and developers are steadily changing their products while capturing all available feedback data [9]. Currently, UES allows adding single feedback entries [4], but so far, an API for integrating in-app feedback in an automated way has been missing.…”