Hitesh Golchha

I am currently pursuing MS in CS from UMass Amherst. I have work and research experience in the fields of NLP, Conversational AI (language generation and understanding), Reinforcement Learning and Representation Learning from companies like Amazon, Flipkart and research labs at UMass Amherst, IIT Patna and Bar Ilan University.

I am currently pursuing research in the IESL Lab under Prof. Andrew McCallum , investigating ways to use LLMs and RL to play Text Based Games.

As an Applied Scientist intern at Amazon last summer, I worked on probing Product representations from vision, text, and multi-modal models for identifying products which have similar export-eligibility.

For two years at Flipkart (Walmart), I worked on problems in Conversational AI: intent detection, entity extraction, sentence segmentation and rescoring ASR beams. Challenges included handling multilinguality, class imbalance, miscalibration, augmentation and robustness. The voice assistants we developed are used by millions of Indians for E-commerce shopping every day. I also participated in hackathons on unsupervised reverse image search and question answering.

I did my B.Tech. in Computer Science and Engineering from Indian Institute of Technology (IIT) Patna, India. I was guided by Prof. Asif Ekbal and Prof. Pushpak Bhattacharyya for my B.Tech Thesis and during my Junior Research Fellowship. My research endeavours were on Style transfer for dialog response generation, Natural Language Understanding for agents, Suggestion Mining, and Question answering.

I also did a short research internship at Bar Ilan University under Prof. Ido Dagan after my Junior Year on Proposition Coreference Resolution for generating Natural Language Knowledge graphs.

June update: I started my YouTube channel, "The Random Variable". I intend to discuss ideas, papers, math and code. Latest video: "Reformer explained (Paper + 🤗Hugging Face code)"

Email  /  CV  /  Google Scholar  /  Twitter  /  Github  /  YouTube  / 

My research has focused on solving several semantic tasks in Natural Langauge Processing using Machine Learning algorithms.

We develop a system that transforms a generic customer care response into a 'Courteous' one. The model takes into context the emotions and content of the conversation history as well while generating and is trained on MLE+RL.

We compare multi-task approaches with standalone and pipeline ones for Spoken Language Understanding tasks on ATIS, TRAINS and FRAMES datasets.

We develop a linguistically informed, self-supervised model to identify customer-to-customer suggestions in Hotel and Electronics review datasets.

– Text-based Games using RL : Using LLMs, RL and Contrastive Learning for playing text-based science games. – Energy based models for Multilabel Classification : Exploring architectures for Energy Based Models for Multilabel Classification which can model contextual label dependencies.

Probing Product representations from vision, text, and multi-modal models to find Products having similar export-eligibility. Also made contributions to other ML projects within the team performing diverse subset sampling.

Multiple research tasks in the NLU: Intent detection, slot filling, multi-intent sentence segmentation, data augmentation, ASR beam rescoring. Challenges included handling multilinguality, class imbalance, miscalibration, augmentation and robustness. The voice assistants we developed are used by millions of Indians for E-commerce shopping every day. I also participated in hackathons on unsupervised reverse image search and question answering.

Research in the domain of conversational agents: NLU and Style Transfer in Response Generation.

Research in Coreference Resolution of Propositions. I got an exposure to several tasks in NLP associated with creation of Natural Language Knowledge Graphs.

Research in Suggestion Mining, Question Answering