Surjya Ghosh Surjya Ghosh


About me

I am a Ph.D. student in the Department of Computer Science and Engineering, IIT Kharagpur, since Jan 2015. My supervisors are Prof. Bivas Mitra and Prof. Niloy Ganguly. I am a part of the Complex Network Research Group (CNeRG) at IIT Kharagpur. In my research, I closely collaborate with Prof. Pradipta De from Georgia Southern University, USA . My research interest lies in the area of Mobile Affective Computing and Human Computer Interaction. Additionally, I also work on the research problems related to transport systems.

 


Research | Publications | Projects | Academics | Experience | Achievements | Resume | Contact


RESEARCH ABSTRACT [Back to top]

I am working in the area of Mobile Affective Computing. I focus on different aspects of affective computing e.g. affect determination, self-report collection, self-reflection using ubiquitous smart mobile devices. Currently, I am focusing to determine affect based on non-intrusive, low overhead information sources present in smartphone and devising intelligent self-report collection methods. For further details please visit the Projects page.

Keyboard Interaction Driven Emotion Detection on Smartphone [ACII'17] [MobileHCI'17] [IJHCS'19]

A significant amount of smartphone interaction involves keyboard interaction (typing, swyping) owing to the large amount of usage of IM applications like WhatsApp, FB messenger. In these applications, people often express their emotions using different emojis. This provides an opportunity to trace the keyboard interaction details along the emotions reported. Moreoever, the feeling of an emotion often persists for some duration before fading. All these together (typing, swyping, emotion persistence) can be leveraged to develop a non-intrusive way of emotion detection from smartphone interaction.

Experience Sampling Method in Emotion Detection [MCSS(UbiComp Adj)'15] [CCNC'17] [TAFFC'19]

One of the key challenges in developing smartphone based emotion detection applications is to collect the emotion self-reports, which are used as emotion ground truth. In most of the emotion detection systems, the self-reports are collected via Experience Sampling Method (ESM). Designing an ESM schedule involves probing the user repeatedly at suitable moments to collect self-reports. Timely probe generation to collect high fidelity user responses while keeping probing rate low is challenging. In mobile-based ESM, timeliness of the probe is also impacted by user's availability to respond to self-report request. Thus, a good ESM design must consider - probing frequency, timely self-report collection, and notifying at opportune moment to ensure high response quality.

Application of Emotion Detection [Mobicom'18] [COMSNETS'19] [IUI'19]

(a) EmoKey: We develop an emotion-aware keyboard EmoKey for monitoring emotion state over long-period of time. In order to make the predictions accurate we deploy an deep neural network (DNN) on the device. We also measure the resource overhead of executing such a DNN model on the device.

(b) Keyboard Layout Optimization: The advances in affective computing have led to the development of many affect-aware applications. As human emotion can be inferred unobtrusivley based on text-input interaction in smartphone, it becomes imperative to optimize the keyboard layout based on human emotion. We investigate the correlation between auto-suggest usage and huamn emotion, which opens the scope for optimizing the layout by dynamically displaying the auto-suggest facility.


PUBLICATIONS [Back to top]

Google Scholar | DBLP

Journal

Conference


ACADEMICS [Back to top]

Education

Teaching Assistantships


PROFESSIONAL EXPERIENCE [Back to top]

Industry Experience


ACHIEVEMENTS [Back to top]


CONTACT ME [Back to top]


Date modified: Mar 27, 2019