Feature-Less End-to-End Nested Term Extraction

Gao, Yuze; Yuan, Yu

doi:10.1007/978-3-030-32236-6_55

Cited by 9 publications

(6 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kucza et al (2018) applied sequence labeling method to extract relevant terms. (Gao and Yuan 2019) proposed a novel term extraction method based on span classification and span ranking on the top of CNN architecture.…”

Section: Related Workmentioning

confidence: 99%

Multilingual Automatic Term Extraction in Low-Resource Domains

Sadat

2021

FLAIRS

View full text Add to dashboard Cite

With the emergence of the neural networks-based approaches, research on information extraction has benefited from large-scale raw texts by leveraging them using pre-trained embeddings and other data augmentation techniques to deal with challenges and issues in Natural Language Processing tasks. In this paper, we propose an approach using sequence-to-sequence neural networks-based models to deal with term extraction for low-resource domain. Our empirical experiments, evaluating on the multilingual ACTER dataset provided in the LREC-TermEval 2020 shared task on automatic term extraction, proved the efficiency of deep learning approach, in the case of low-data settings, for the automatic term extraction task.

show abstract

Section: Related Workmentioning

confidence: 99%

Multilingual Automatic Term Extraction in Low-Resource Domains

Sadat

2021

FLAIRS

View full text Add to dashboard Cite

show abstract

“…However, in recent years, research into ATE has outgrown this linguistic/statistical/hybrid typology. Gao and Yuan (2019) propose a typology of five, calling the three original categories "rule-based", "statistical", and "hybrid", and adding "machine-learning based" and "deep-learning based". While it is true that the original typology is due for an update, their suggestion may not be ideal, in the sense that "deep learning" is technically a type of "machine learning", which is considered the opposite of "rule-based".…”

Section: Related Researchmentioning

confidence: 99%

“…Most approaches fit into one of these three categories relatively easily, though there may be some exceptions. For instance, Gao and Yuan (2019) use a sequential approach with deep learning and, rather than traditional IOBlabelling, they work with all possible term spans in each sentence, with spans up to a maximum term length k, where k must be smaller than or equal to the sentence length. This allows the detection of nested terms with a sequential approach.…”

Section: Related Researchmentioning

confidence: 99%

“…The final category is devoted to features that use word-or character-embeddings, which are becoming ever more prevalent. Recently, embeddings are used in both featurebased and so-called "featureless" methodologies (Gao and Yuan 2019;Wang, Liu, and McDonald 2016). In the TermFrame project (Pollak et al 2019) FastText embeddings trained on the small, domain-specific corpus are used to extend the list of CTs obtained through a traditional, hybrid approach.…”

Section: Related Researchmentioning

confidence: 99%

See 1 more Smart Citation

Hamlet

Terryn

Hoste

Lefever

2021

TERM

View full text Add to dashboard Cite

Automatic term extraction (ATE) is an important task within natural language processing, both separately, and as a preprocessing step for other tasks. In recent years, research has moved far beyond the traditional hybrid approach where candidate terms are extracted based on part-of-speech patterns and filtered and sorted with statistical termhood and unithood measures. While there has been an explosion of different types of features and algorithms, including machine learning methodologies, some of the fundamental problems remain unsolved, such as the ambiguous nature of the concept “term”. This has been a hurdle in the creation of data for ATE, meaning that datasets for both training and testing are scarce, and system evaluations are often limited and rarely cover multiple languages and domains. The ACTER Annotated Corpora for Term Extraction Research contain manual term annotations in four domains and three languages and have been used to investigate a supervised machine learning approach for ATE, using a binary random forest classifier with multiple types of features. The resulting system (HAMLET Hybrid Adaptable Machine Learning approach to Extract Terminology) provides detailed insights into its strengths and weaknesses. It highlights a certain unpredictability as an important drawback of machine learning methodologies, but also shows how the system appears to have learnt a robust definition of terms, producing results that are state-of-the-art, and contain few errors that are not (part of) terms in any way. Both the amount and the relevance of the training data have a substantial effect on results, and by varying the training data, it appears to be possible to adapt the system to various desired outputs, e.g., different types of terms. While certain issues remain difficult – such as the extraction of rare terms and multiword terms – this study shows how supervised machine learning is a promising methodology for ATE.

show abstract

“…Logistic Regression model LIBLINEAR in [12] was proposed to classify if a word is actually a term or not, investigating different types of features. Reference [13] proposed a method called end-to-end nested term extraction using both classification and ranking to distinguish whether terms are conceptual for a specific domain. They use both CNN and LSTM neural networks to improve their results.…”

Section: Background and Related Workmentioning

confidence: 99%

An Unsupervised Semantic Model for Arabic/French Terminology Extraction

Setha

Aliane

2021

Proceedings of International Conference on Emerging Technologies and Intelligent Systems

View full text Add to dashboard Cite

We provide in this work, a new approach for automatic bilingual terminology extraction involving Arabic and French languages. We aim to build a bilingual list of terms for a specific field, namely "Algerian Culture". The approach is based on building a semantic space for context understanding, adopting COALS (Correlated Occurrence Analogue to Lexical Semantics). COALS is an interesting semantic model which combines the advantages of LSA and HAL models. We use "Wikipedia" to build a comparable corpus for "Algerian heritage and culture" and a translation tool for term alignment.

show abstract

Feature-Less End-to-End Nested Term Extraction

Cited by 9 publications

References 15 publications

Multilingual Automatic Term Extraction in Low-Resource Domains

Multilingual Automatic Term Extraction in Low-Resource Domains

Hamlet

An Unsupervised Semantic Model for Arabic/French Terminology Extraction

Contact Info

Product

Resources

About