An app for collecting smartphone sensor data labeled with student activities
DataCollection is an app for the collection of smartphone sensor data while carrying out typical student activities.
You can download the app here. The app requires an Android 4.0 or newer.
At the first use you are asked to enter an e-mail adress. This is only used to contact you in the case you won a price. Among all participants collecting recordings for at least ten different activities (min. 10 minutes samples each) we lot five vouchers worth 30 € from a local book store (Habel). We associate your e-mail adress to the data only to check whether you collected the amount of recordings to participate in the price draw. Afterwards the association between the recording and your e-mail adress is deleted.
Collecting Sensor Data
Any time you start one of the following activities you are asked to collect sensor data. Start your app, select the activity you will perform following from the pull down menu an press the record button. The collection takes place still yout stop it. During the collection put the phone at the place you usually keep it, e.g. in your pocket or on the table. Please don't leave the phone in your bag though, since this significantly affects the quality of the data. The app shows you the duration of the current collection in minutes.
After performing the activity stop the recording.
We collect the following data:
If you want, you can opt out of uploading your audio data. In this case we will extract some features from the audio files and upload them instead. These features cannot be used to reconstruct the original sound.
Uploading labeled Sensor Data
After the collection of some labled data you are asked to upload the collected files to our server. We suggest to do it only if you have access to W-LAN. For uploading the data select from the menu "Upload Data" and click "Upload". After the preparation oft the data the upload starts. The app shows you the upload process. When done the app informs you "Upload done".
If you interrupt the upload the data are stored until you start the next upload. After a complete upload all sensor data are deleted from your phone.
If you realize you forgot to stop the recording and thus got something personal recorded, no worry! Just click the "Delete" button (scroll under the delete button), and the last recording will be deleted.
We are working on new applications to support learning in different scenarios. In the current project we plan to develop a new learning diary. In contrast to a usual diary where the user is asked to take down his or her activities, we aim to do it automatically. Therefore we are working on algorithms which detect the learners' activities automatically.
What would this mean for you? You could have an app you install at the beginning of the semester, which logs all your learning activities and their duration. Then, periodically or as often as you like, you can get an overview of your activities and reflect on how to improve your learning. (Did attending lectures mean you needed less time to learn at home? Does doing exercises help you get a better grade?)
But first we need your help!
We need sensor data to train our algorithms, in order to be able to recognize learning activities. So you can use our app and record only in the time period you decide.
If you are interested in our work we refer to the following publications:
Irina Diaconita, Christoph Rensing, Stephan Tittel:
Getting the Information You Need, When You Need It: A Context-aware Q&A System for Collaborative Learning. In: Christoph Rensing, Sara de Freitas, Tobias Ley, Pedro J. Muñoz-Merino: Open Learning and Teaching in Educational Communities. Proceedings of the 9th European Conference on Technology Enhanced Learning (Lecture Notes in Computer Science), vol. 8719, p. 410-415, Springer International Publishing, September 2014.
Irina Diaconita, Andreas Reinhardt, Delphine Christin, Christoph Rensing:
Bleep Bleep! Determining Smartphone Locations by Opportunistically Recording Notification Sounds. In: Proceedings of the International Conference on Mobile and Ubiquitous Systems: Computing, Networking, and Services, p. 110-119 , ICST (Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering) , December 2014.
In case you have any questions or comments please contact us: