Browsing by Author "Sahami Shirazi, Alireza"

Now showing 1 - 1 of 1

Open Access
Exploiting & sharing context : computer mediated nonverbal communication
(2014) Sahami Shirazi, Alireza; Schmidt, Albrecht (Prof. Dr.)
Humans are social beings and need to communicate and share their emotions. Communication takes place by exchanging not only verbal information but also nonverbal information. With the development of human civilization, communication is undergoing a constant change. The advances in technologies have led to new communication mediums allowing non-colocated persons to communicate and exchange information. Further, the ubiquity of computers, such as mobile phones, has provided the possibility to use computing technologies for different means in various contexts. Users most often carry the devices with themselves and are even emotionally attached to them. Such computer-mediated communication is generally non face-to-face and communicators are in different contexts. While face-to-face communication consists of verbal and nonverbal information, the lack of nonverbal and contextual information in non-face-to-face communication prevents effective communication and may lead to confusion. Therefore, exploiting and sharing contextual information is essential to enhance the communication between non-colocated persons. This thesis investigates how to exploit physiological and cognitive information to retrieve awareness about users themselves and their contexts as well as sharing such information using nonverbal modalities through computer-mediated communication channels. It discusses how information about certain user's activities can be obtained using brain signals and user's explicit interactions. Further, it explores nonverbal modalities as a communication channel to express and share context and awareness. The research questions are addressed using empirical methods commonly applied in the human-computer interaction research domain. In the initial step, we explore two sources as means to obtain the user's context and monitor specific activities. We, first, assess brain signals acquired from commercial brain-computer interfaces (BCI) to determine common activities, i.e., reading, listening, and relaxing. We further assess how the user's emotional state correlates with emotional information provided by the BCIs using videos as stimuli. Second, we investigate how only explicit interactions with mobile applications, instead of using any sensor, can be used to determine the user's physical activities. In particular, we explore how the explicit interaction can be utilized to monitor sleeping as one of the prime everyday activities. Monitoring sleep information shows not only one's daily routines but also indicates the physical state. We assess how exchanging information about one's sleep behavior impacts behavior and awareness in communication. We conduct user studies in the controlled setups and in the wild using application stores to obtain findings with high internal and external validity. In the next step, we investigate sharing context information using nonverbal channels. We explore rhythm-based tones as a nonverbal mean for communication. We assess how melody composition can be used as a way to express and share emotions. Music, in general, can communicate one's state of mind and it is often characterized as the language of emotion. We use short messages on mobile phones, as one of the most popular services on mobile phones at this time, for sharing emotions. Furthermore, we examine how audio previewing of messages can be used to communicate contents and enable awareness. The current notification approaches such as visual cues and audio tones aim at solely informing the receiver that a message arrived without revealing any further information. We propose an algorithm for audio previewing messages in such a way that content and intention of text messages is additionally conveyed. In the final step, we explore iconic interfaces on mobile phones as a nonverbal modality for sharing sentiments and connect non-colocated users. Through a use case, we assess how the sentiments collected via this channel correlates with moments in a real-time while watching TV. We carry out a study with a large number of users to assess this approach in a realistic context. The experience gained while conducting several studies in the wild using the application stores allowed us to identify challenges and limitations of this methodology. Further, reviewing prior work that used similar approaches enabled us to have a comprehensive overview about advantages and disadvantages of such studies. Based on the findings, we propose best practices how such user studies can be carried out. We discuss aspects and challenges that should be taken into account during designing such user studies. The contributions of this thesis provide insights into using physiological and cognitive data to determine activities and emotional states of users and obtain context information. We present how explicit interactions with a mobile application can be leveraged to monitor sleeping behavior of users without using any wearable sensor. It further presents that rhythm-based tones and iconic user interfaces, as nonverbal modalities, can be used to share contextual information. We discuss how sharing context information can affect users awareness and connectedness. The practices for research through the applications stores can be used as a guideline for researchers who want to address their research questions through this research methodology.