Joy Bose: Projects
Responsive and Adaptive Web BrowsingJun 2013
We investigate systems to make the web browser adaptive and responsive to the needs of the user.
Publications
Contextual Adaptive User Interface For Android Devices. In Proc. INDICON, Nov 2013, IEEE
Responsive, Adaptive and User Personalized Rendering on Mobile Browsers. in Proc. International Workshop on Software Engineering for Web Applications (SEWAD), Aug 2014, IEEE
Rendering on Browsers Responsive to User Head Position and Environmental Conditions. In Proc INDICON, Dec 2014, IEEE
Humanitarian Engineering
Jun 2012
Investigating methods in which technology can be made accessible to people in rural or remote areas
Publications
A solution for a mobile computing device along with supporting infrastructure for the needs of illiterate users in rural areas. In Proc. India Conference (INDICON) 2012, IEEE
A kiosk based model for employment generation in rural areas, published in IEEE Global Humanitarian technology conference South Asian Satellite (GHTC-SAS) in Kovalam, September 2014, IEEE
Encrypting sensor data based on user gestures
Jun 2012 - Jun 2013
We present methods to encrypt and decrypt data in mobile devices such as SMS messages, based on user gestures.
Publications
Encryption in mobile devices using sensors. in Proc. Sensors Applications Symposium (SAS), Galveston, USA, 2013 IEEE.
We present methods to encrypt and decrypt data in mobile devices such as SMS messages, based on user gestures. Publications Encryption in mobile devices using sensors. in Proc. Sensors Applications Symposium (SAS), Galveston, USA, 2013 IEEE.
Bias Detection in web browsers
Jan 2016 - Jun 2018
We investigate methods to detect bias in news articles being read on the mobile web browser in real time.
Publications:
Bias Based Navigation for News Articles and Media, in Proc. NLDB 2016
Paper on Bias Aware Web Browsing selected in ICMLA 2018.
Increasing the click rate of web notifications
Jan 2015 - Oct 2016
We present a system to increase the click rate of push notifications, by predicting which categories of notifications the user is more likely to click. For doing this, we build a model that takes the user's past clicks and online behavior as inputs, and outputs the time of day when the user is more likely to click a notification belonging to a particular category. E.g. sports on friday evenings.
Publications:
A Unified Push Notifications Service for Mobile Devices. In Proc. CONECCT 2015, IIIT Bangalore, India. IEEE.
Secure Web Push System. In Proc. COMSNETS 2016, Bangalore, India. IEEE.
Seamless Push Service with Flow Control for Embedded Devices. in Proc. CCNC 2016, Las Vegas, USA, IEEE.
Intelligent Web Push Architecture with Push Flow Control and Push Continuity. In Proc. 23rd IEEE International Conference on Web Services, June 2016, San Francisco, USA. IEEE
US Patent US 20160014057 A1. Method and system for providing dynamically customized web push messages in a wireless network. Filed July 2015. https://www.google.com/patents/US20160014057
Modelling attention in Web Browsers using EEG
Jan 2014 - Jan 2016
Project Goal: To explore uses of attention (as detected and measured by portable EEG sensors such as NeuroSky MindWave) while browsing the web and performing other activities on smartphones
Publications:
METHOD AND SYSTEM FOR GENERATING GRAPHICAL TRANSCRIPT OF A USER DURING AN INTERACTIVE SESSION: India Patent 6328/CHE/2014, Filed Dec 2014
EEG Based Detection of Area of Interest in a Web Page (ICACCI 2015 Conference)
A web browser responsive to the user interest level (INDICON 2015 Conference)
A Hands Free Browser Using EEG and Voice Inputs (IJCNN 2015 Conference)
Attention Sensitive Web Browsing (ACM Compute 2016 Conference)
Awards: Best Paper Award, Track 4, Indicon 2015
Calendar event classification
Jun 2017 - Aug 2018
Built machine learning models to classify calendar data input (short text data) by users into one of predefined categories, in both Korean and English, using an ensemble of machine learning techniques.
Publications
A Personalized Health Recommendation System based on Smartphone Calendar Events (ICOST 2018)
System to extract relevant content on a webpage for reader mode
Aug 2018 - Apr 2019
The aim of this project was to develop a machine learning application to extract content from a webpage based on relevance to the reader and remove ads or boilerplate content. We tried different machine learning algorithms and used a customized version of an internal virtual DOM extractor tool.
Publications
Extraction of relevant images for boilerplate removal in web browsers. INDICON 2019
Machine learning to predict field labels in an autofill system
Apr - Dec 2019
Built a machine learning model to predict the field labels for the autofill feature in web browsers. Used WinML and saved the model in ONNX format.
Determining the optimal nudges for improving a student's score in standardized tests
Jan - Jun 2020
We used optimization algorithms to identify the optimal parameters where a student could improve their score by a certain amount, based on the historical data of the student. This is used to generate appropriate nudge messages to help the student improve the test scores. The student score in tests using the online learning platform is a function of multiple parameters related to test taking behavior and knowledge in different concepts.
Publications
Auto generation of diagnostic assessments and their quality evaluation. Published in Educational Data Mining Conference (EDM 2020) in Morocco
Identifying categories of financial variance for a telecom provider
2020-2021
Worked on a project to classify financial variance data into one of a number of categories using machine learning. The objective was to help save time of telecom providers in a previously manual task for root cause identification for investigating variance in financial data.