Usage scenarios. 38
4.2.1. Motivation and Goal 38
4.2.2. Approach 39 4.2.3. Usage Scenarios 39
Usage scenarios. The usage scenarios of SAFE are manifold. Perhaps the most intuitive scenario for the app feature extraction is monitoring app features health, which is to continuously identify and measure which app features are mentioned in user comments, how frequent, and their associated sentiments. In addition, app store analytics approaches that classify user reviews into bug reports and feature requests or into informative and non-informative can benefit from SAFE. If, for example, a bug report gets enriched by the addressed app feature, the information can help developers to faster trace and solve the bug. Another scenario SAFE provides is the identification of the features delta and new insights. This can be done by identifying the differences between features described by developers and those described in the user reviews. If the delta is identified, one might want to extract further details by distinguishing between different reviews subpopulations into specific regions, stores, or channels like a forum or social media comments. Those could be recommended to stakeholder to turn them into actual requirements. In the paper, we address the different use of language. Developers write their text more formal and often try to emphasize the app features. Users on the other hand often write colloquial, use abbreviations, emoticons, and put strong emotions in their text. Developers and users might learn how to communicate better and bridge their “vocabulary gap”. On the other hand, developers can easily get the list of suggested features by users, which is a useful information for requirements identification, scoping, and prioritization. Current research classifies user reviews in e.g., feature requests and bug reports. But research often simply classifies the complete app review but does not filter the relevant information. With SAFE we might be able to also enable the clustering and aggregation of artifacts (such as comments) based on features.
Usage scenarios. The OpenReq trial partners have large corpora of user comments that provide an important resource for them. Our approach might be employed to a) structure software users’ comments;
b) filter and synthesize rationale-backed comments for certain stakeholders, such as developers, analysts, or other users; c) improve communication among stakeholders. In particular, user rationale classifiers might be employed to structure user comments by highlighting polarizations and the broad spectrum of rationale and perspectives in their discussions. This would enable stakeholders to quickly survey the growing number of user feedback and to help users to easily provide feedback or to get involved into existing discussions. User rationale classifiers can be also used for visualization of pro- and contra- users’ stances which highlight contrasting user perceptions on, for example, non-functional requirements (such as pro and contra stances on software usability). For community managers, user rationale can be used to mark, filter, and synthesize potentially low/highly informative comments. Those, for instance, that have a low justification density can be filtered out. Furthermore, rationale-backed comments of special interest can be selected and explored—e.g., comments mentioning conclusive decisions on switching to a new software product. This allows stakeholders to inspect the best user rationale concepts which can be highlighted or searched for within the comments. User rationale classifiers can help in requirements prioritization and negotiation. For instance, an issue can be highly prioritized when it is frequently mentioned along with a reason for abandoning a software product. Alternatives can give insights on how software features are used together and which compatibility requirements should be considered. The broad spectrum of their justifications can help to get insights and to better understand users’ different perspectives (e.g., on usability). In particular, the whole communication among stakeholders can be improved—e.g., by reusing users’ arguments during requirements negotiation. Furthermore, users’ justifications can enrich existing requirements documentation and design decisions.
Usage scenarios. To create a capacity management usage scenario:
Usage scenarios. NCF supports two different types of messaging: - Publish-Subscribe. In this type of message pattern there is a publisher that creates messages and there are zero, one or more listeners that subscribe to those messages. The publisher is not aware of whom is listening. Note that it can be set up in such a way that only one listener will receive the messages.
