Technically, this results in a very large number of notifications. The app achieves this by generating a new notification each second until the call is ended. When a user is voice chatting using the Skype app, for example, the app shows and updates the current duration of the ongoing call in the notification bar. Apps use this feature based on their internal logic. In the Android OS design guideline it says: “ Think of noti- user is not paying attention ”. One of the interesting aspects in the dataset is the number of notifications that the different apps show a user per day. In the following we use these categories for further analysis. Table 1 shows the derived categories and provides an overview of the number of users, the apps in the respective category, the amount of click time data points, and the total number of ratings we collected through our questionnaire. These new categories then were discussed and finalized, which resulted in 14 categories total. To achieve this, two researchers independently reviewed the selected apps and each derived new categories. This encour- aged us to develop a more specific categorization scheme. Beside, prior work categorization is not publicly available or do not include all apps in our dataset. However, previous work shows that they are used differently. For example, email apps (e.g., Gmail or Outlook), text messaging apps (e.g., Whatsapp or SMS apps), and Voice messaging apps (e.g., Google Hang- out and Skype) were all in the same Communication category. Further, we realized the categories retrieved from the Google Play marketplace were too generic. This resulted in 173 apps from 23 categories (70.9% of the total number of notifications). In the first step we only included apps with at least 400 users in a new dataset. Hence, we selected a subset of the data for the analysis. A large number of apps in the dataset were only used by very few users. The users used 20,014 unique apps from 30 different categories that we retrieved from the Google Play marketplace. The three most frequent locales are en US (29.7%), es ES (10.35%), and en GB (9.85%). In total 19.06% of the notifications stored in the database were polled by the plug-in and shown on the desktop computer. Between the 10th of January 2013 and the 19th of July 2013 we collected 197,515,366 notifications from 40,191 unique users (253 unique locales) using Desktop Notification. Uploaded content of notifications is automatically and periodically deleted to preserve users’ privacy and anonymity. To avoid ethical issues and preserve the users’ privacy we did not collect any information about the content of notifications due to the high sensitivity of the information that can be included in notifications. The text that accompanies notifications can include very sen- sitive information, including SMS bodies, email titles, and the user’s contacts. collected the apps added to the blacklist from the users.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |