Question Answering

Question answering is concerned with answer user questions, either in a closed-domain or open-domain setting automatically. We are interested in exploiting synergies between question answering and related tasks, such as framing entailment or relation extraction as question answering tasks. Further research interests are question answering in conversational settings, such as for chatbots.

We were prevously involved in two funded projects related to this theme: a project on subjectivity in question answering, funded by a faculty research award from Megagon Labs; as well a PhD fellowship grant on conversational question answering for customer support from BotXO, which Ana Valeria Gonzalez works on.

nlu question-answering

Publications

Question Answering (QA) datasets have been instrumental in developing and evaluating Large Language Model (LLM) capabilities. However, …
PDF Project Project

Answering complex queries on incomplete knowledge graphs is a challenging task where a model needs to answer complex logical queries in …
PDF Project Project Project

There have been many efforts to try to understand what grammatical knowledge (e.g., ability to understand the part of speech of a …
PDF Project Project

Two of the most fundamental challenges in Natural Language Understanding (NLU) at present are: (a) how to establish whether deep …
PDF Project Project

Alongside huge volumes of research on deep learning models in NLP in the recent years, there has been also much work on benchmark …
PDF Project

Subjectivity is the expression of internal opinions or beliefs which cannot be objectively observed or verified, and has been shown to …
PDF Dataset Project Project

Learning what to share between tasks has been a topic of high importance recently, as strategic sharing of knowledge has been shown to …
PDF Code Project Project Project Project

It is challenging to automatically evaluate the answer of a QA model at inference time. Although many models provide confidence scores, …
PDF Project

Task oriented dialogue systems rely heavily on specialized dialogue state tracking (DST) modules for dynamically predicting user intent …
PDF Project Project

Task oriented dialogue systems rely heavily on specialized dialogue state tracking (DST) modules for dynamically predicting user intent …
PDF Project

Although the vast majority of knowledge bases KBs are heavily biased towards English, Wikipedias do cover very different topics in …
PDF Project Project Project Project

The best systems at the SemEval-16 and SemEval-17 community question answering shared tasks – a task that amounts to question …
PDF Project Video

Many Machine Reading and Natural Language Understanding tasks require reading supporting text in order to answer questions. For …
PDF Code Project