This annual award launched by IKDD in 2022, would recognize the best doctoral dissertation(s) in the broad areas of Data Science, Artificial Intelligence, Machine Learning, Data Mining, Knowledge Discovery and applications of Data-driven techniques. The IKDD Doctoral Dissertation Award winner and up to two runners-up will be recognized at the annual CODS-COMAD conference, and their dissertations will have the opportunity to be published on the IKDD Web site (https://ikdd.acm.org/). The award winner will receive a plaque and a cheque for INR 1,00,000. The runners-up will receive a plaque and a cheque of INR 50,000. The winner and runners-up will also receive a free registration to attend the CODS-COMAD conference and will be invited to present their work in a special session at the conference.
Nomination Deadline: July 31, 2022 Aug 7, 2022
Nominated candidates must be from a PhD granting institute in India. Further, the nominated candidate should have successfully defended his/her thesis between August 1, 2021 and July 31, 2022 (i.e., in the one-year period before the deadline for application). Every PhD granting institute in India will be allowed to nominate 1 candidate from across all disciplines working in relevant areas. Institutes that produce more than 10 dissertations in relevant areas during the one year period under consideration can nominate 2 students. Institutes should self-declare in case they are nominating 2 students and furnish supporting evidence.
Nominations must be made by the doctoral advisor. All nomination materials must be in English and in PDF format. Late submissions will not be accepted. A nomination must include:
The dissertation can be nominated in parallel for other national and global awards, such as the ACM, the ACM India and the SIGKDD Doctoral Dissertation awards.
All nomination materials must be submitted using this Easy Chair link. All required documents must be uploaded as a zip file.
Winner
Thesis: Rational Deep Machines: Towards Explainable, Trustworthy and Robust Deep Learning Systems
Institute: Indian Institute of Technology, Hyderabad
IKDD congratulates Anirban Sarkar for making impactful contributions to explainability and robustness of deep neural networks models from multiple perspectives including post hoc interpretability, ante hoc interpretability, causality, and attributional robustness - some of which have been translated to practical use.
Runner-Up
Thesis: Encode-Attend-Refine-Decode: Enriching Contextual Representations for Natural Language Generation
Institute: Indian Institute of Technology, Madras
IKDD congratulates Preksha Nema for her contributions to natural language generation by introducing models which avoid repetitions, provide better coverage of input information and use task specific reward functions
Runner-Up
Thesis: Compression for Distributed Optimization and Timely Updates
Institute: Indian Institute of Science, Bengaluru
IKDD congratulates Prathamesh Mayekar for his work on developing information-theoretically optimal quantization techniques for gradient compression in stochastic optimization with new algorithms and theoretical lower bounds to establish optimality.