Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Student Research Workshop

26 papers
Computationally Efficient Wasserstein Loss for Structured Labels (2021.eacl-srw)

Copied to clipboard

Challenge: Existing approaches to estimate the probability distribution of labels are based on tree-Wasserstein distance.
Approach: They propose a tree-Wasserstein distance regularized LDL algorithm for hierarchical text classification tasks.
Outcome: The proposed method performs well on synthetic and real-world datasets and compares favorably with the Sinkhorn algorithm in terms of computation time and memory usage.
Have Attention Heads in BERT Learned Constituency Grammar? (2021.eacl-srw)

Copied to clipboard

Challenge: Recent pre-trained language models have gained great success in many tasks, but what they have learned, and when they perform well remain unknown.
Approach: They employ the syntactic distance method to extract implicit constituency grammar from attention weights of attention heads of BERT and RoBERTa.
Outcome: The proposed models induce some grammar types much better than baselines, suggesting some heads act as a proxy for constituency grammar.
Do we read what we hear? Modeling orthographic influences on spoken word recognition (2021.eacl-srw)

Copied to clipboard

Challenge: Existing theories and models of spoken word recognition focus on accessing lexical knowledge given an acoustic realization of a word form.
Approach: They propose two models that instantiate hypotheses regarding the influence of orthography on spoken word recognition.
Outcome: The proposed models reproduce human-like behavior in different ways and provide testable hypotheses for future research on the source of orthographic effects in spoken word recognition.
PENELOPIE: Enabling Open Information Extraction for the Greek Language through Machine Translation (2021.eacl-srw)

Copied to clipboard

Challenge: a new method for information extraction from Greek corpora is being developed for low-resource languages.
Approach: They propose a methodology that aims at bridging the gap between high and low-resource languages in the context of Open Information Extraction.
Outcome: The proposed method outperforms the current state-of-the-art for the Greek language on benchmark datasets.
A Computational Analysis of Vagueness in Revisions of Instructional Texts (2021.eacl-srw)

Copied to clipboard

Challenge: We analyze edits that involve cases of vagueness in instructional texts . we extract and analyze version pairs of an instruction before and after a revision .
Approach: They propose to extract and analyze edits that involve cases of vagueness in instructions . they adopt a pairwise ranking task to show improvements over existing baselines .
Outcome: The proposed model can distinguish between two versions of an instruction in a noisy dataset.
A reproduction of Apple’s bi-directional LSTM models for language identification in short strings (2021.eacl-srw)

Copied to clipboard

Challenge: Language Identification is the task of identifying a document’s language.
Approach: They propose to use bi-LSTMs to identify language on very short strings such as text message fragments to perform automatic spell check.
Outcome: The proposed model outperforms open-source language identifiers and its language identification mistakes are due to confusion between related languages.
Automatically Cataloging Scholarly Articles using Library of Congress Subject Headings (2021.eacl-srw)

Copied to clipboard

Challenge: Currently, nearly 40 institutions have registered their repositories with RAMP . manual cataloging of articles using LCSH is a challenge due to the rapid growth of articles .
Approach: They propose to automatically annotate articles with Library of Congress Subject Headings . they use web scraping to extract keywords for a collection of articles from RAMP .
Outcome: The proposed approach predicts LCSH for scholarly articles using keywords extracted from RAMP . the proposed model is validated by a multi-label classification problem.
Model Agnostic Answer Reranking System for Adversarial Question Answering (2021.eacl-srw)

Copied to clipboard

Challenge: Existing methods for adversarial QA are often model specific and require retraining of the model . authors propose a simple method that can be applied directly to any QA model based on a model-agnostic approach .
Approach: They propose a model-agnostic approach that explicitly reranks candidate answers . they use a QA model that scores candidates on the basis of content overlap with the question .
Outcome: The proposed method outperforms state-of-the-art models on adversarial examples without retraining.
BERT meets Cranfield: Uncovering the Properties of Full Ranking on Fully Labeled Data (2021.eacl-srw)

Copied to clipboard

Challenge: Existing information retrieval models based on pre-trained BERT models have been tested on data collections with partial relevance labels, where a relevant document has not been exposed to the annotators.
Approach: They propose to use BERT-based rankers to evaluate documents with partial relevance labels on a Cranfield collection, which comes with full relevance judgment on all documents in the collection.
Outcome: The proposed model performs better than the initial ranker and re-ranker on the Cranfield dataset.
Siamese Neural Networks for Detecting Complementary Products (2021.eacl-srw)

Copied to clipboard

Challenge: e-commerce websites are often too overwhelming for users to find what they need at one place . a recommendation system can detect complementary products using only the purchase history .
Approach: They propose a content-based recommender system for detecting complementary products using Siamese Neural Networks (SNN).
Outcome: The proposed system detects complementary products with 85% accuracy using only the product titles.
Contrasting distinct structured views to learn sentence embeddings (2021.eacl-srw)

Copied to clipboard

Challenge: Existing methods to build sentence embeddings rely on a similar Recurrent Neural Network (RNN) heterogeneity of performances across models and tasks makes us assume some structures might be better adapted given the considered task or sentence.
Approach: They propose a self-supervised method that builds sentence embeddings from syntactic structures . they hypothesize that some linguistic representations might be better adapted given the task .
Outcome: The proposed method outperforms comparable methods on several tasks from standard sentence embedding benchmarks.
Discrete Reasoning Templates for Natural Language Understanding (2021.eacl-srw)

Copied to clipboard

Challenge: Existing approaches to reasoning over multiple parts of a passage provide little evidence of their reasoning process, especially with regards to why specific operands are chosen for a reasoning task.
Approach: They propose a method that decomposes complex questions into subquestions that can take advantage of single-span extraction models and derives the final answer according to instructions in a predefined reasoning template.
Outcome: The proposed approach is interpretable and requires little supervision while competing with the state-of-the-art models.
Multilingual Email Zoning (2021.eacl-srw)

Copied to clipboard

Challenge: Existing literature on email zoning is mainly limited to English . however, it is possible to discern a level of formal organization in the way most emails are formed.
Approach: They propose a multilingual email zoning benchmark based on a language agnostic sentence encoder and a new model that uses a biLSTM with a CRF to classify each sentence into an email zone.
Outcome: The proposed model is competitive with current English benchmarks and reached state-of-the-art performance in English.
Familiar words but strange voices: Modelling the influence of speech variability on word recognition (2021.eacl-srw)

Copied to clipboard

Challenge: Despite the lack of acoustic-phonetic invariance in speech, listeners can reliably recognize spoken words despite the lack aural-phonemic invariancy.
Approach: They propose a deep neural model which is trained to retrieve the meaning of a word given its spoken form, a task which resembles that faced by a human listener.
Outcome: The proposed model is more sensitive to dialectical variation than gender variation and more related to related languages.
Emoji-Based Transfer Learning for Sentiment Tasks (2021.eacl-srw)

Copied to clipboard

Challenge: Sentiment tasks such as hate speech detection and sentiment analysis are often low-resource . a transfer learning approach is used to transfer the emotional information encoded in emojis to a sentiment task .
Approach: They exploit emotional information encoded in emojis to enhance performance on sentiment tasks . they use a transfer learning approach where parameters learned by an e-based source task are transferred to a sentiment target task .
Outcome: The proposed method improves sentiment tasks on languages other than English with high emoji content and label distribution under three conditions.
A Little Pretraining Goes a Long Way: A Case Study on Dependency Parsing Task for Low-resource Morphologically Rich Languages (2021.eacl-srw)

Copied to clipboard

Challenge: Neural dependency parsing has been a success for many domains and languages, but the bottleneck of massive labelled data limits its effectiveness for low resource languages.
Approach: They propose to use morphological knowledge to improve dependency parsing for morphology rich languages in a low-resource setting to perform experiments.
Outcome: The proposed method achieves an average gain of 2 points (UAS) and 3.6 points (LAS) on 10 MRLs in low-resource settings.
Development of Conversational AI for Sleep Coaching Programme (2021.eacl-srw)

Copied to clipboard

Challenge: Existing methods to treat insomnia neglect conversational aspects, which plays a critical role in sleep therapy.
Approach: They propose to develop conversational AI for a sleep coaching programme which is motivated by CBT-I treatment and provide an automated analytic system to support human experts.
Outcome: The proposed system could interact naturally with a user and provide an automated analytic system to support human experts.
Relating Relations: Meta-Relation Extraction from Online Health Forum Posts (2021.eacl-srw)

Copied to clipboard

Challenge: Relation extraction is a key task in knowledge extraction, and is often defined as identifying relations that hold between entities in text.
Approach: They propose to conceptualise relation extraction tasks for user-generated health texts and create a dataset and model for meta-relation extraction.
Outcome: The proposed model will be able to extract meta-relations from user-generated health texts with tolerable cognitive load and a new dataset and annotation scheme with tolerance for annotations.
Towards Personalised and Document-level Machine Translation of Dialogue (2021.eacl-srw)

Copied to clipboard

Challenge: State-of-the-art (SOTA) neural machine translation systems translate texts at sentence level, ignoring context.
Approach: They propose to integrate extra-textual information into the translation process for the domain of dialogue extracted from TV subtitles in five languages: English, Brazilian Portuguese, German, French and Polish.
Outcome: The proposed systems translate texts at sentence level, ignoring context . there are no readily available robust evaluation metrics for them .
Semantic-aware transformation of short texts using word embeddings: An application in the Food Computing domain (2021.eacl-srw)

Copied to clipboard

Challenge: Recent work in food computing focus on generating new recipes from scratch . however, there are a large number of new recipes generated daily with user reviews .
Approach: They propose to use word embedding models to capture the semantic meaning of recipe ingredients and use them to enrich their data.
Outcome: The proposed engine will use food data to modify a recipe to fit user preferences.
TMR: Evaluating NER Recall on Tough Mentions (2021.eacl-srw)

Copied to clipboard

Challenge: a NER evaluation tool is available via a repository.
Approach: They propose to use Tough Mentions Recall to supplement traditional named entity recognition evaluation by examining recall on specific subsets of ”tough” mentions.
Outcome: The proposed metrics enable differentiation between otherwise similar-scoring systems and identify patterns in performance that would go unnoticed from overall precision, recall, and F1.
The Effectiveness of Morphology-aware Segmentation in Low-Resource Neural Machine Translation (2021.eacl-srw)

Copied to clipboard

Challenge: Current NMT systems typically operate at the level of subwords, causing problems of vocabulary sparsity.
Approach: They compare subword segmentation methods with morphologically-based methods in a low-resource setting . they find that no consistent and reliable differences emerge between the methods .
Outcome: The proposed methods outperform BPE in a low-resource translation setting.
Making Use of Latent Space in Language GANs for Generating Diverse Text without Pre-training (2021.eacl-srw)

Copied to clipboard

Challenge: Existing models for generating diverse texts are not pre-trained . generative adversarial networks suffer from mode-collapsing if they are not trained .
Approach: They propose a GAN model that produces diverse texts conditioned by latent code . they propose to use Gumbel-Softmax distribution for word sampling .
Outcome: The proposed model is competitive with existing models, which requires pre-training.
Beyond the English Web: Zero-Shot Cross-Lingual and Lightweight Monolingual Classification of Registers (2021.eacl-srw)

Copied to clipboard

Challenge: Existing studies on register classification for web documents have limited results due to skewed datasets and low performance.
Approach: They propose two new register-annotated corpora for French and Swedish . they show that deep pre-trained language models perform strongly in these languages .
Outcome: The proposed models outperform existing models in English and Finnish and can match or surpass existing models.
Explaining and Improving BERT Performance on Lexical Semantic Change Detection (2021.eacl-srw)

Copied to clipboard

Challenge: Lexical semantic change detection is still a challenging field due to the success of type-based embeddings in SemEval-2020 Task 1 and other NLP tasks.
Approach: They compare the performance of BERT embeddings with results from the word sense disambiguation dataset underlying SemEval-2020 Task 1 and the Italian follow-up task DIACR-Ita.
Outcome: The proposed model outperforms token-based embeddings on lexical semantic change detection tasks.
Why Find the Right One? (2021.eacl-srw)

Copied to clipboard

Challenge: a new study investigates the impact of anaphoric one words in English on the neural machine translation process.
Approach: They investigate the impact of anaphoric one words in English on the Neural Machine Translation process using English-Hindi as source and target language pair.
Outcome: The proposed system performs poorly on sentences containing anaphoric ones compared to sentences involving regular, non-anaphorical ones . the results show that amongst the anamorphic words, the noun class is clearly much harder for NMT than the determinatives .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations