Data

Tools

Indexes

Applications

Experiments

Open Source Science

adatalks

https://github.com/clips/adatalks

Science Score: 28.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ○
    codemeta.json file
  • ○
    .zenodo.json file
  • ○
    DOI references
  • ✓
    Academic publication links
    Links to: arxiv.org
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (6.9%) to scientific vocabulary
Last synced: 10 months ago · JSON representation ·

Repository

Basic Info
  • Host: GitHub
  • Owner: clips
  • Language: TeX
  • Default Branch: master
  • Size: 335 KB
Statistics
  • Stars: 2
  • Watchers: 9
  • Forks: 1
  • Open Issues: 1
  • Releases: 0
Created almost 6 years ago · Last pushed almost 3 years ago
Metadata Files
Readme Citation

README.md

ADA TaLKS

Our work on Conversational Agents can be described as Adaptive Dialogue Agents using Task-oriented Learning with Knowledge and Senses.

ADA TaLKS represents the CLiPS research part of the Flanders AI Research Program focusing on conversational agents as part of the human-like AI challenge. Within our work package on conversational agents, we cooperate with the Universities of Leuven (LIIR), Brussels (AI-LAB) and Ghent (LT3 and K2T). On this page you can find short video's on all the current research in our workpackage.

The focus of ADA TaLKS is on enabling a flexible method for incorporating and updating relevant background knowledge in a deep learning approach for dialog systems. In addition, we are researching memory encoding in order to improve consistency of the generated answers with the dialog history and the agent's persona.

People

Publications

The Effects of Expressing Empathy/Autonomy Support Using a COVID-19 Vaccination Chatbot: Experimental Study in a Sample of Belgian Adults
Wojciech Trzebiński, Toni Claessens, Jeska Buhmann, Aurélie De Waele, Greet Hendrickx, Pierre Van Damme, Walter Daelemans, Karolien Poels
JMIR Form Res 2023
bibtex

20Q: Overlap-Free World Knowledge Benchmark for Language Models
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
GEM 2022
bibtex

Is It Smaller Than a Tennis Ball? Language Models Play the Game of Twenty Questions
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
BlackboxNLP 2022
bibtex

Machine Translation for Multilingual Intent Detection and Slots Filling
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
MMNLU 2022
bibtex

Domain- and Task-Adaptation for VaccinChatNL, a Dutch COVID-19 FAQ Answering Corpus and Classification Model
Jeska Buhmann, Maxime De Bruyn, Ehsan Lotfi and Walter Daelemans
COLING 2022
bibtex | Outstanding paper

Open-Domain Dialog Evaluation Using Follow-Ups Likelihood
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
COLING 2022
bibtex

Teach Me What to Say and I Will Learn What to Pick: Unsupervised Knowledge Selection Through Response Generation with Pretrained Generative Models
Ehsan Lotfi, Maxime De Bruyn, Jeska Buhmann and Walter Daelemans
NLP4ConvAI 2021
bibtex | Best Paper Award

MFAQ: a Multilingual FAQ Dataset
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
MRQA 2021
bibtex | Best Paper Award

ConveRT for FAQ Answering
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
Bnaic/benelearn 2021
bibtex

BART for Knowledge Grounded Conversations
Maxime De Bruyn, Ehsan Lotfi, Jeska Buhmann and Walter Daelemans
KDD Converse 2020
bibtex

Proofs of Technology

VaccinChat

We developed a Dutch chatbot that answers FAQ's on COVID-19 vaccines and the Flemish vaccination strategy: vaccinchat.be

This project also gave us insight in the evolution of user questions over time

Car insurance chatbot for Belfius

In this work different data augmentation methods were tested to lift the performance of Belfius’ less-used French car insurance bot to the level of the more frequently used Dutch system. These methods include back-translation, paraphrasing and question generation.

Owner

  • Name: Computational Linguistics Research Group
  • Login: clips
  • Kind: organization
  • Location: CLiPS

Computational Linguistics and Psycholinguistics Research Center, University of Antwerp

Citation (citations/2021.mrqa-1.1.bib) 

@inproceedings{de-bruyn-etal-2021-mfaq,
    title = "{MFAQ}: a Multilingual {FAQ} Dataset",
    author = "De Bruyn, Maxime  and
      Lotfi, Ehsan  and
      Buhmann, Jeska  and
      Daelemans, Walter",
    booktitle = "Proceedings of the 3rd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2021",
    address = "Punta Cana, Dominican Republic",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.mrqa-1.1",
    doi = "10.18653/v1/2021.mrqa-1.1",
    pages = "1--13",
    abstract = "In this paper, we present the first multilingual FAQ dataset publicly available. We collected around 6M FAQ pairs from the web, in 21 different languages. Although this is significantly larger than existing FAQ retrieval datasets, it comes with its own challenges: duplication of content and uneven distribution of topics. We adopt a similar setup as Dense Passage Retrieval (DPR) and test various bi-encoders on this dataset. Our experiments reveal that a multilingual model based on XLM-RoBERTa achieves the best results, except for English. Lower resources languages seem to learn from one another as a multilingual model achieves a higher MRR than language-specific ones. Our qualitative analysis reveals the brittleness of the model on simple word changes. We publicly release our dataset, model, and training script.",
}

GitHub Events

Total
Last Year