In terms of data flow, any data transferred from a Smartphone client, say an acceptance of invitation to join a group, would be sent immediately to the server, which would store the content in the database along with any relevant organizational information (group membership changes) and metadata (system usage timestamps). Relevant outbound data, such as notification of the new member joining the group, is then routed to the appropriate message queue for distribution. If the recipient is an SMS user, the outbound message is sent immediately once it gets to the front of the SMS outgoing message queue. Due to battery drainage issues, the Smartphone clients check the web service for new data every 10 minutes. If there are new data, or if the user has sent a message, the Smartphone clients checks for new content every 30 seconds until there is no activity, at which point it returns to a 10 minute data check cycle. Smartphone users can initiate a send/receive if desired. Once the new data arrives, the Smartphone client integrates it into a local store for faster display in the user interface and offline usage.
Figure 5: Slam system architecture schematic
3. EXPERIMENTAL Field STUDY
An experimental field study was conducted with Slam to assess the potential social benefits of group-based mobile messaging. Groups of participants used the Slam prototype as described above for group-wide messaging as well as a slightly modified version of Slam that permitted only one-to-one messaging. Groups used each version of Slam for 7 - 10 days. Usage behavior was logged and participants completed questionnaires throughout the study, as well as provided qualitative feedback in a round-table wrap-up discussion. In addition to the impact of group-based messaging on social connections, secondary questions targeted the role of different usage modalities (e.g., coordination, social sharing).