1 code implementation • COLING 2022 • Israa Alghanmi, Luis Espinosa-Anke, Steven Schockaert
Interpreting patient case descriptions has emerged as a challenging problem for biomedical NLP, where the aim is typically to predict diagnoses, to recommended treatments, or to answer questions about cases more generally.
no code implementations • LREC 2022 • Yixiao Wang, Zied Bouraoui, Luis Espinosa Anke, Steven Schockaert
Many applications crucially rely on the availability of high-quality word vectors.
no code implementations • LREC 2022 • Carla Perez Almendros, Luis Espinosa Anke, Steven Schockaert
Patronizing and Condescending Language (PCL) is a subtle but harmful type of discourse, yet the task of recognizing PCL remains under-studied by the NLP community.
no code implementations • SemEval (NAACL) 2022 • Carla Perez-Almendros, Luis Espinosa-Anke, Steven Schockaert
This paper presents an overview of Task 4 at SemEval-2022, which was focused on detecting Patronizing and Condescending Language (PCL) towards vulnerable communities.
no code implementations • EMNLP (WNUT) 2020 • Israa Alghanmi, Luis Espinosa Anke, Steven Schockaert
A particularly striking example is the performance of AraBERT, an LM for the Arabic language, which is successful in categorizing social media posts in Arabic dialects, despite only having been trained on Modern Standard Arabic.
1 code implementation • EMNLP 2021 • Asahi Ushio, Jose Camacho-Collados, Steven Schockaert
Among others, this makes it possible to distill high-quality word vectors from pre-trained language models.
no code implementations • 25 Mar 2024 • Hanane Kteich, Na Li, Usashi Chatterjee, Zied Bouraoui, Steven Schockaert
We show that this leads to embeddings which capture a more diverse range of commonsense properties, and consistently improves results in downstream tasks such as ultra-fine entity typing and ontology completion.
no code implementations • 25 Mar 2024 • Na Li, Thomas Bailleux, Zied Bouraoui, Steven Schockaert
One line of work treats this task as a Natural Language Inference (NLI) problem, thus relying on the knowledge captured by language models to identify the missing knowledge.
no code implementations • 23 Feb 2024 • Nitesh Kumar, Usashi Chatterjee, Steven Schockaert
We focus in particular on the task of ranking entities according to a given conceptual space dimension.
no code implementations • 29 Jan 2024 • Victor Charpenay, Steven Schockaert
Region based knowledge graph embeddings represent relations as geometric regions.
no code implementations • 18 Dec 2023 • Frank Mtumbuka, Steven Schockaert
Relation extraction is essentially a text classification problem, which can be tackled by fine-tuning a pre-trained language model (LM).
no code implementations • 23 Oct 2023 • Amit Gajbhiye, Zied Bouraoui, Na Li, Usashi Chatterjee, Luis Espinosa Anke, Steven Schockaert
We show that by augmenting the label set with shared properties, we can improve the performance of the state-of-the-art models for this task.
1 code implementation • 18 Oct 2023 • Nitesh Kumar, Steven Schockaert
A common strategy is to rely on knowledge graphs (KGs) such as ConceptNet, and to model the relation between two concepts as a set of paths.
no code implementations • 9 Oct 2023 • Usashi Chatterjee, Amit Gajbhiye, Steven Schockaert
The theory of Conceptual Spaces is an influential cognitive-linguistic framework for representing the meaning of concepts.
1 code implementation • 30 Sep 2023 • Asahi Ushio, Jose Camacho-Collados, Steven Schockaert
In particular, we show that masked language models such as RoBERTa can be straightforwardly fine-tuned for this purpose, using only a small amount of training data.
1 code implementation • 26 Sep 2023 • Shahul ES, Jithin James, Luis Espinosa-Anke, Steven Schockaert
We introduce RAGAs (Retrieval Augmented Generation Assessment), a framework for reference-free evaluation of Retrieval Augmented Generation (RAG) pipelines.
1 code implementation • 14 Aug 2023 • Akash Anil, Víctor Gutiérrez-Basulto, Yazmín Ibañéz-García, Steven Schockaert
The task of inductive knowledge graph completion requires models to learn inference patterns from a training graph, which can then be used to make predictions on a disjoint test graph.
no code implementations • 24 May 2023 • Asahi Ushio, Jose Camacho Collados, Steven Schockaert
Relations such as "is influenced by", "is known for" or "is a competitor of" are inherently graded: we can rank entity pairs based on how well they satisfy these relations, but it is hard to draw a line between those pairs that satisfy them and those that do not.
no code implementations • 22 May 2023 • Na Li, Zied Bouraoui, Steven Schockaert
In this paper, we show that the performance of existing methods can be improved using a simple technique: we use pre-trained label embeddings to cluster the labels into semantic domains and then treat these domains as additional types.
1 code implementation • 22 May 2023 • Frank Mtumbuka, Steven Schockaert
In this paper, we propose to improve on this process by pre-training an entity encoder such that embeddings of coreferring entities are more similar to each other than to the embeddings of other entities.
1 code implementation • 16 May 2023 • Na Li, Hanane Kteich, Zied Bouraoui, Steven Schockaert
Second, concept embeddings should capture the semantic properties of concepts, whereas contextualised word vectors are also affected by other factors.
no code implementations • 15 May 2023 • Simone Tedeschi, Johan Bos, Thierry Declerck, Jan Hajic, Daniel Hershcovich, Eduard H. Hovy, Alexander Koller, Simon Krek, Steven Schockaert, Rico Sennrich, Ekaterina Shutova, Roberto Navigli
In the last five years, there has been a significant focus in Natural Language Processing (NLP) on developing larger Pretrained Language Models (PLMs) and introducing benchmarks such as SuperGLUE and SQuAD to measure their abilities in language understanding, reasoning, and reading comprehension.
no code implementations • 11 Oct 2022 • Steven Schockaert
In particular, we find that when the epistemic pooling principle is satisfied, in most cases it is impossible to verify the satisfaction of propositional formulas using linear scoring functions, with two exceptions: (i) max-pooling with embeddings that are upper-bounded and (ii) Hadamard pooling with non-negative embeddings.
1 code implementation • COLING 2022 • Amit Gajbhiye, Luis Espinosa-Anke, Steven Schockaert
Grasping the commonsense properties of everyday concepts is an important prerequisite to language understanding.
no code implementations • 2 Dec 2021 • Kun Yan, Chenbin Zhang, Jun Hou, Ping Wang, Zied Bouraoui, Shoaib Jameel, Steven Schockaert
A key feature of the multi-label setting is that images often have multiple labels, which typically refer to different regions of the image.
1 code implementation • 21 Sep 2021 • Asahi Ushio, Jose Camacho-Collados, Steven Schockaert
Among others, this makes it possible to distill high-quality word vectors from pre-trained language models.
no code implementations • AKBC 2021 • Steven Schockaert
However, little is known about what kinds of semantic dependencies can be modelled in this way.
1 code implementation • ACL (RepL4NLP) 2021 • Yixiao Wang, Zied Bouraoui, Luis Espinosa Anke, Steven Schockaert
Second, rather than learning a word vector directly, we use a topic model to partition the contexts in which words appear, and then learn different topic-specific vectors for each word.
1 code implementation • Findings (ACL) 2021 • Israa Alghanmi, Luis Espinosa-Anke, Steven Schockaert
Pre-trained language models such as ClinicalBERT have achieved impressive results on tasks such as medical Natural Language Inference.
no code implementations • 21 May 2021 • Kun Yan, Zied Bouraoui, Ping Wang, Shoaib Jameel, Steven Schockaert
While the use of class names has already been explored in previous work, our approach differs in two key aspects.
1 code implementation • ACL 2021 • Asahi Ushio, Luis Espinosa-Anke, Steven Schockaert, Jose Camacho-Collados
Analogies play a central role in human commonsense reasoning.
no code implementations • 10 May 2021 • Steven Schockaert, Yazmín Ibáñez-García, Víctor Gutiérrez-Basulto
Ontologies formalise how the concepts from a given domain are interrelated.
no code implementations • 1 Feb 2021 • Kun Yan, Zied Bouraoui, Ping Wang, Shoaib Jameel, Steven Schockaert
The aim of few-shot learning (FSL) is to learn how to recognize image categories from a small number of training examples.
no code implementations • 4 Dec 2020 • Na Li, Zied Bouraoui, Jose Camacho Collados, Luis Espinosa-Anke, Qing Gu, Steven Schockaert
While the success of pre-trained language models has largely eliminated the need for high-quality static word vectors in many NLP applications, such vectors continue to play an important role in tasks where words need to be modelled in the absence of linguistic context.
no code implementations • SEMEVAL 2020 • Shelan Jeawak, Luis Espinosa-Anke, Steven Schockaert
We describe the system submitted to SemEval-2020 Task 6, Subtask 1.
1 code implementation • COLING 2020 • Rana Alshaikh, Zied Bouraoui, Shelan Jeawak, Steven Schockaert
This is exploited by an associated gating network, which uses pre-trained word vectors to encourage the properties that are modelled by a given embedding to be semantically coherent, i. e. to encourage each of the individual embeddings to capture a meaningful facet.
no code implementations • COLING 2020 • Carla Pérez-Almendros, Luis Espinosa-Anke, Steven Schockaert
In this paper, we introduce a new annotated dataset which is aimed at supporting the development of NLP models to identify and categorize language that is patronizing or condescending towards vulnerable communities (e. g. refugees, homeless people, poor families).
no code implementations • 25 Jun 2020 • Yazmín Ibáñez-García, Víctor Gutiérrez-Basulto, Steven Schockaert
In this paper, we instead propose an inductive inference mechanism which is based on a clear model-theoretic semantics, and can thus be tightly integrated with standard deductive reasoning.
no code implementations • 13 Dec 2019 • Zied Bouraoui, Antoine Cornuéjols, Thierry Denœux, Sébastien Destercke, Didier Dubois, Romain Guillaume, João Marques-Silva, Jérôme Mengin, Henri Prade, Steven Schockaert, Mathieu Serrurier, Christel Vrain
Some common concerns are identified and discussed such as the types of used representation, the roles of knowledge and data, the lack or the excess of information, or the need for explanations and causal understanding.
no code implementations • 3 Dec 2019 • Zied Bouraoui, Jose Camacho-Collados, Luis Espinosa-Anke, Steven Schockaert
Unfortunately, meaningful regions can be difficult to estimate, especially since we often have few examples of individuals that belong to a given category.
no code implementations • 28 Nov 2019 • Zied Bouraoui, Jose Camacho-Collados, Steven Schockaert
Starting from a few seed instances of a given relation, we first use a large text corpus to find sentences that are likely to express this relation.
no code implementations • CONLL 2019 • Rana Alshaikh, Zied Bouraoui, Steven Schockaert
To address this gap, we analyze how, and to what extent, a given vector space embedding can be decomposed into meaningful facets in an unsupervised fashion.
no code implementations • 16 Oct 2019 • Yerai Doval, Jose Camacho-Collados, Luis Espinosa-Anke, Steven Schockaert
While monolingual word embeddings encode information about words in the context of a particular language, cross-lingual embeddings define a multilingual space where word embeddings from two or more languages are integrated together.
Cross-Lingual Natural Language Inference Cross-Lingual Word Embeddings +3
1 code implementation • WS 2019 • Yilun Zhou, Julie A. Shah, Steven Schockaert
Commonsense procedural knowledge is important for AI agents and robots that operate in a human environment.
no code implementations • LREC 2020 • Yerai Doval, Jose Camacho-Collados, Luis Espinosa-Anke, Steven Schockaert
Cross-lingual word embeddings are vector representations of words in different languages where words with similar meaning are represented by similar vectors, regardless of the language.
1 code implementation • ACL 2019 • Luis Espinosa Anke, Steven Schockaert, Leo Wanner
Lexical relation classification is the task of predicting whether a certain relation holds between a given pair of words.
no code implementations • ACL 2019 • Shoaib Jameel, Steven Schockaert
To this end, our model relies on the assumption that context word vectors are drawn from a mixture of von Mises-Fisher (vMF) distributions, where the parameters of this mixture distribution are jointly optimized with the word vectors.
1 code implementation • ACL 2019 • Jose Camacho-Collados, Luis Espinosa-Anke, Steven Schockaert
While word embeddings have been shown to implicitly encode various forms of attributional knowledge, the extent to which they capture relational information is far more limited.
no code implementations • SEMEVAL 2019 • Carla P{\'e}rez-Almendros, Luis Espinosa-Anke, Steven Schockaert
This paper summarizes our contribution to the Hyperpartisan News Detection task in SemEval 2019.
1 code implementation • 17 May 2019 • Jose Camacho-Collados, Yerai Doval, Eugenio Martínez-Cámara, Luis Espinosa-Anke, Francesco Barbieri, Steven Schockaert
Cross-lingual embeddings represent the meaning of words from different languages in the same vector space.
1 code implementation • 21 Feb 2019 • Yilun Zhou, Steven Schockaert, Julie A. Shah
In this paper we instead propose to learn to predict path quality from crowdsourced human assessments.
no code implementations • EMNLP 2018 • Francesco Barbieri, Luis Espinosa-Anke, Jose Camacho-Collados, Steven Schockaert, Horacio Saggion
Human language has evolved towards newer forms of communication such as social media, where emojis (i. e., ideograms bearing a visual meaning) play a key role.
1 code implementation • CONLL 2018 • Thomas Ager, Ond{\v{r}}ej Ku{\v{z}}elka, Steven Schockaert
In this paper, we argue that there is an inherent trade-off between capturing similarity and faithfully modelling features as directions.
1 code implementation • EMNLP 2018 • Yerai Doval, Jose Camacho-Collados, Luis Espinosa-Anke, Steven Schockaert
Cross-lingual word embeddings are becoming increasingly important in multilingual NLP.
1 code implementation • COLING 2018 • Luis Espinosa-Anke, Steven Schockaert
For example, by examining clusters of relation vectors, we observe that relational similarities can be identified at a more abstract level than with traditional word vector differences.
no code implementations • COLING 2018 • Zied Bouraoui, Shoaib Jameel, Steven Schockaert
Given a set of instances of some relation, the relation induction task is to predict which other word pairs are likely to be related in the same way.
no code implementations • COLING 2018 • Steven Schockaert
Conceptual spaces, as proposed by Grdenfors, are similar to entity embeddings, but provide more structure.
no code implementations • ACL 2018 • Shoaib Jameel, Zied Bouraoui, Steven Schockaert
Word embedding models such as GloVe rely on co-occurrence statistics to learn vector representations of word meaning.
no code implementations • NAACL 2018 • Luis Espinosa-Anke, Steven Schockaert
Automatically identifying definitional knowledge in text corpora (Definition Extraction or DE) is an important task with direct applications in, among others, Automatic Glossary Generation, Taxonomy Learning, Question Answering and Semantic Search.
no code implementations • 26 May 2018 • Víctor Gutiérrez-Basulto, Steven Schockaert
To address this shortcoming, in this paper we introduce a general framework based on a view of relations as regions, which allows us to study the compatibility between ontological knowledge and different types of vector space embeddings.
no code implementations • 3 May 2018 • Zied Bouraoui, Steven Schockaert
Several recently proposed methods aim to learn conceptual space representations from large text collections.
no code implementations • 17 Apr 2018 • Ondrej Kuzelka, Yuyi Wang, Steven Schockaert
In many applications of relational learning, the available data can be seen as a sample from a larger relational structure (e. g. we may be given a small fragment from some social network).
no code implementations • 15 Mar 2018 • Ondrej Kuzelka, Yuyi Wang, Jesse Davis, Steven Schockaert
We consider the problem of predicting plausible missing facts in relational data, given a set of imperfect logical rules.
no code implementations • 14 Nov 2017 • Shoaib Jameel, Zied Bouraoui, Steven Schockaert
Word embedding models such as GloVe rely on co-occurrence statistics from a large corpus to learn vector representations of word meaning.
no code implementations • 5 Oct 2017 • Gustav Sourek, Martin Svatos, Filip Zelezny, Steven Schockaert, Ondrej Kuzelka
Lifted Relational Neural Networks (LRNNs) describe relational domains using weighted first-order rules which act as templates for constructing feed-forward neural networks.
no code implementations • 18 Sep 2017 • Ondrej Kuzelka, Yuyi Wang, Jesse Davis, Steven Schockaert
In the propositional setting, the marginal problem is to find a (maximum-entropy) distribution that has some given marginals.
no code implementations • 21 Aug 2017 • Zied Bouraoui, Shoaib Jameel, Steven Schockaert
Word embeddings have been found to capture a surprisingly rich amount of syntactic and semantic knowledge.
no code implementations • CONLL 2017 • Shoaib Jameel, Steven Schockaert
Although region representations of word meaning offer a natural alternative to word vectors, only few methods have been proposed that can effectively learn word regions.
no code implementations • 21 Jun 2017 • Bei Shi, Wai Lam, Shoaib Jameel, Steven Schockaert, Kwun Ping Lai
Word embedding models such as Skip-gram learn a vector-space representation for each word, based on the local word collocation patterns that are observed in a text corpus.
no code implementations • 19 May 2017 • Ondrej Kuzelka, Jesse Davis, Steven Schockaert
Compared to Markov Logic Networks (MLNs), our method is faster and produces considerably more interpretable models.
no code implementations • COLING 2016 • Shoaib Jameel, Steven Schockaert
We propose a new word embedding model, inspired by GloVe, which is formulated as a feasible least squares optimization problem.
no code implementations • 18 Nov 2016 • Ondrej Kuzelka, Jesse Davis, Steven Schockaert
In this paper, we advocate the use of stratified logical theories for representing probabilistic models.
no code implementations • 18 Apr 2016 • Ondrej Kuzelka, Jesse Davis, Steven Schockaert
We introduce a setting for learning possibilistic logic theories from defaults of the form "if alpha then typically beta".
no code implementations • 18 Feb 2016 • Shoaib Jameel, Steven Schockaert
Conceptual spaces are geometric representations of conceptual knowledge, in which entities correspond to points, natural properties correspond to convex regions, and the dimensions of the space correspond to salient features.
no code implementations • 16 Dec 2015 • Sofie De Clercq, Steven Schockaert, Martine De Cock, Ann Nowé
Since the introduction of the stable marriage problem (SMP) by Gale and Shapley (1962), several variants and extensions have been investigated.
1 code implementation • 3 Jun 2015 • Ondrej Kuzelka, Jesse Davis, Steven Schockaert
Markov logic uses weighted formulas to compactly encode a probability distribution over possible worlds.
no code implementations • 9 Oct 2014 • Steven Schockaert, Sanjiang Li
First, we identify all ways in which the set of RCC8 base relations can be restricted to guarantee that consistent networks can be convexly realized in respectively 1D, 2D, 3D, and 4D.
no code implementations • 30 Nov 2013 • Kim Bauters, Steven Schockaert, Martine De Cock, Dirk Vermeir
In particular, while the complexity of most reasoning tasks coincides with standard disjunctive ASP, we find that brave reasoning for programs with weak disjunctions is easier.
no code implementations • 28 Feb 2013 • Sofie De Clercq, Steven Schockaert, Martine De Cock, Ann Nowé
Our encoding can easily be extended and adapted to the needs of specific applications.