Mohnish Dubey

• Lead to the development of question answering system over DBpedia.
• Here Natural language questions are normalized into an intermediary canonical syntactic form and then translated into a SPARQL
• A rule based system for question answering over DBpedia Knowledge Graph
• source code https://github.com/AskNowQA/AskNowNQS

• Lead the project with aim to create a Question Answering data set with the pair of Natural Language question to their corresponding SPARQL query.
• LC-QuAD 1.0 published a dataset of 5,000 question over DBpedia.
• LC-QuAD 2.0 published a dataset of 30,000 question over Wikidata.
• project website: http://lc-quad.sda.tech

• Developed the component to performs entity linking and relation linking as a joint single task.
• First by modeling the problem to Generalised Travelling Salesman Problem.
• Second using machine learning in order to exploit the connection density between nodes in the knowledge graph.
• Source code: https://github.com/AskNowQA/EARL

• In this internal Fraunhofer project we developed a template based question answering system
• We targeted ARD Video annotations, and fetch videos based on user’s input in German language.

• A large scale dataset for answer verbaliser for simple natural questions.
• The answer in this dataset are syntactically and semantically closer to the question than to the KG triple.
• Website https://sda.tech/projects/vanilla/

• Orchestrating the development of RAG-LLM-based chatbot system designed for nurses. Overseeing aspects such as Total Cost of Ownership, Synthetic Data Generation, and implementing Guardrails within the chatbot to ensure safety and reliability
• Knowledge Acquisition and KG construction for Recommendation System aimed at the continuous earning and development of (MR) Technicians.
• Embraced the dual role of Scrum Master, rigorously applying lean agile principles to guide the team through sprint ceremonies, optimizing workflow for peak performance

• Developing use-cases and user stories for the dialogue system with the vision of the project of active healthy ageing
• Investigating several components(such as translation services) and pipelines to use them to meet the project requirements
• Developing a Dialogue system with RASA framework and further integrating Question Answering capabilities.

• Defended Ph.D. Thesis on Question Answering over Knowledge Graph (KGQA). (January 2021)
• Published state of the art systems on KGQA, Entity and Relation Linking and Dataset generation.
• Managed and supervised student assistants for KGQA project, AskNow.
• Conducted research on Answer Representation and managed UI development for KGQA.
• Conducted NLP Labs and Seminars, currently also mentoring master student projects

• Worked on Semantic Question Answering over Knowledge Graph.
• Lead developer in projects on Question Answering over Video Annotations in German language.
• Supported with evaluation on Dialogue System for Car Infotainment.
• Generated Question Answering benchmarking for Conversation AI systems in specific domains such as COVID-19 data, and Financial reports.

• In the six month internship worked and gained knowledge in the field of Semantic Web and Question Answering.
• Learned about Knowledge gGraphs and SPARQL Query.
• Worked on rule based translation of English Questions to SPARQL query.

During my PhD at the University of Bonn, I was a part of the Smart Data Analytics group lead by Prof. Jens Lehmann.

At Technology Dhirubhai Ambani Institute of Information & Communication Technology, I completed my master’s degree with machine intelligence speciation. In my master thesis, I worked on the normalization of negative sentences. Here I was first introduced to Semantic Web by Dr Sourish Dasgupta.

I completed my Engineering degree in Computer Science at Jai Narain College of Technology affiliated to Rajiv Gandhi Proudyogiki Vishwavidyalaya, India. During my last year, I scored 98 percentile in the GATE exam, motivating me to study computer science further.