Timezone: »

Inductive Quantum Embedding
Santosh Kumar Srivastava · Dinesh Khandelwal · Dhiraj Madan · Dinesh Garg · Hima Karanam · L Venkata Subramaniam

Tue Dec 08 09:00 AM -- 11:00 AM (PST) @ Poster Session 1 #386

Quantum logic inspired embedding (aka Quantum Embedding (QE)) of a Knowledge-Base (KB) was proposed recently by Garg:2019. It is claimed that the QE preserves the logical structure of the input KB given in the form of unary and binary predicates hierarchy. Such structure preservation allows one to perform Boolean logic style deductive reasoning directly over these embedding vectors. The original QE idea, however, is limited to the transductive (not inductive) setting. Moreover, the original QE scheme runs quite slow on real applications involving millions of entities. This paper alleviates both of these key limitations. We start by reformulating the original QE problem to allow for the induction. On the way, we also underscore some interesting analytic and geometric properties of the solution and leverage them to design a faster training scheme. As an application, we show that one can achieve state-of-the-art performance on the well-known NLP task of fine-grained entity type classification by using the inductive QE approach. Our training runs 9-times faster than the original QE scheme on this task.

Author Information

Santosh Kumar Srivastava (IBM Research AI)
Dinesh Khandelwal (IBM Research AI)

I am a Research Scientist in AI Reasoning group working at IBM Research Lab, Delhi, India. I have completed Ph.D. in Machine Learning from IIT Delhi. My primary research interests lie in the area of Deep Learning, Probabilistic Graphical Models, and Question Answering. I also holds a Master's degree in Machine Learning from IISc Bangalore.

Dhiraj Madan (IBM Research)
Dinesh Garg (IBM Research AI, India)
Hima Karanam (IBM Research AI)
L Venkata Subramaniam (IBM Research AI - India)

More from the Same Authors