With the most advanced natural language processing and artificial intelligence approaches, effective summarization of long and multi-topic documents---such as academic papers---for readers from different domains still remains a challenge. To address this, we introduce ConceptEVA, a mixed-initiative approach to generate, evaluate, and customize summaries for long and multi-topic documents. ConceptEVA incorporates a custom multi-task longformer encoder decoder to summarize longer documents. Interactive visualizations of document concepts as a network reflecting both semantic relatedness and co-occurrence help users focus on concepts of interest. The user can select these concepts and automatically update the summary to emphasize them. We present two iterations of ConceptEVA evaluated through an expert review and a within-subjects study. We find that participants' satisfaction with customized summaries through ConceptEVA is higher than their own manually-generated summary, while incorporating critique into the summaries proved challenging. Based on our findings, we make recommendations for designing summarization systems incorporating mixed-initiative interactions.
We present ChameleonControl, a real-human teleoperation system for scalable remote instruction in hands-on classrooms. In contrast to existing video or AR/VR-based remote hands-on education, ChameleonControl uses a real human as a surrogate of a remote instructor. Building on existing human-based telepresence approaches, we contribute a novel method to teleoperate a human surrogate through synchronized mixed reality hand gestural navigation and verbal communication. By overlaying the remote instructor's virtual hands in the local user's MR view, the remote instructor can guide and control the local user as if they were physically present. This allows the local user/surrogate to synchronize their hand movements and gestures with the remote instructor, effectively teleoperating a real human. We deploy and evaluate our system in classrooms of physiotherapy training, as well as other application domains such as mechanical assembly, sign language and cooking lessons. The study results confirm that our approach can increase engagement and the sense of co-presence, showing potential for the future of remote hands-on classrooms.
Collaborative cloud platforms make it easier and more convenient for multiple users to work together on files (GoogleDocs, Office365) and store and share them (Dropbox, OneDrive). However, this can lead to privacy and security conflicts between the users involved, for instance when a user adds someone to a shared folder or changes its permissions. Such multiuser conflicts (MCs), though known to happen in the literature, have not yet been studied in-depth. In this paper, we report a study with 1,050 participants about MCs they experienced in the cloud. We show what are the MCs that arise when multiple users work together in the cloud and how and why they arise, what is the prevalence and severity of MCs, what are their consequences on users, and how do users work around MCs. We derive recommendations for designing mechanisms to help users avoid, mitigate, and resolve MCs in the cloud.