We have developed question answering systems under different scenarios.
For Machine Reading Comprehension (MRC), we proposed different methods to improve the accuracy of answering questions over passages [
AAAI'20;
ACL'23].
For Knowledge Graph Question Answering (KGQA), we investegated and improved different modules in the QA framework, which achieve SOTA results in a variety of scenarios (e.g., simple questions [
IJCAI'19], complex questions [
WSDM'20;
ACL'20;
ACL'23], conversational questions [
ACL'21]).
For Question Answering with Databases, we explored to generate complex questions in a low-resource condition [
EMNLP'23] and align Large Language Models (LLMs) to a domain-specific database [
CIKM'24].
For Multi-modal Question Answering, we discussed the applications of LLMs and its safety issue [
MM'23;
IJCAI'24;
ECCV'24].
There are surveys [
TKDE'22] you can start with.
We have demonstration pages [
Demonstration Page] that you can try on!
