Hasso-Plattner-InstitutSDG am HPI
Hasso-Plattner-InstitutDSG am HPI

Connecting Text and Knowledge using Deep Learning

Artificial Intelligence and Intelligent Systems Group
Hasso Plattner Institute

Office: G-3.2.08
Tel.: + 49- (0) 331 5509-4926
Email: manoj.prabhakar(at)hpi.de
Links: Homepage LinkedIn

Member of the HPI Research School since November 2020

Supervisor: Prof. Dr. Gerard de Melo

Research Overview

My research deals with connecting text and knowledge using deep learning methods. We focus on information extraction, knowledge graphs and investigate them using various interesting approaches. Currently, I am working on a couple of projects related to this idea.


  • Fact Comparison

Facts are an important aspect of the web. Some of the popular knowledge bases like DBpedia, Freebase, Wikidata, etc. contain a lot of facts and there are many possible applications using them. In this research, we investigate methods to compare facts in a knowledge graph and analyze these facts using various approaches like combining graph embeddings and large language models.

  • Fact Extraction

A fundamental problem in information extraction is to extract relevant information like entities and their relations from unstructured text. This can be split into two seperate tasks: named entity recognition and relation extraction. We are working on the analysis of various existing approaches and also developing new efficient methods.


Manoj Prabhakar Kannan Ravi, Kuldeep Singh, Isaiah Onando Mulang', Saeedeh Shekarpour, Johannes Hoffart, and Jens Lehmann : CHOLAN: A Modular Approach for Neural Entity Linking on Wikipedia and Wikidata. Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume(EACL), 2021.


As Teaching Assistant / Tutor

As Advisor

  • Bias in Knowledge Graphs (MSc) (Summer 2021-present)
  • Commit Message Generation using Transformers (MSc) (Winter 2020/21-present)