Automatic Term Extraction Research Articles

A Knowledge-Driven Approach for Automatic Semantic Aspect Term Extraction Using the Semantic Power of Linked Open Data

Aspect-Based Sentiment Analysis (ABSA) is a crucial process for assessing customer feedback and gauging satisfaction with products or services. It typically consists of three stages: Aspect Term Extraction (ATE), Aspect Categorization Extraction (ACE), and Sentiment Analysis (SA). Various techniques have been proposed for ATE, including unsupervised, supervised, and hybrid methods. However, many studies face challenges in detecting aspect terms due to reliance on training data, which may not cover all multiple aspect terms and relate semantic aspect terms effectively. This study presents a knowledge-driven approach to automatic semantic aspect term extraction from customer feedback using Linked Open Data (LOD) to enrich aspect extraction outcomes in the training dataset. Additionally, it utilizes the N-gram model to capture complex text patterns and relationships, facilitating accurate classification and analysis of multiple-word terms for each aspect. To assess the effectiveness of the proposed model, experiments were conducted on three benchmark datasets: SemEval 2014, 2015, and 2016. Comparative evaluations with contemporary unsupervised, supervised, and hybrid methods on these datasets yielded F-measures of 0.80, 0.76, and 0.77, respectively.

Can cross-domain term extraction benefit from cross-lingual transfer and nested term labeling?

Automatic term extraction (ATE) is a natural language processing task that eases the effort of manually identifying terms from domain-specific corpora by providing a list of candidate terms. In this paper, we treat ATE as a sequence-labeling task and explore the efficacy of XLMR in evaluating cross-lingual and multilingual learning against monolingual learning in the cross-domain ATE context. Additionally, we introduce NOBI, a novel annotation mechanism enabling the labeling of single-word nested terms. Our experiments are conducted on the ACTER corpus, encompassing four domains and three languages (English, French, and Dutch), as well as the RSDO5 Slovenian corpus, encompassing four additional domains. Results indicate that cross-lingual and multilingual models outperform monolingual settings, showcasing improved F1-scores for all languages within the ACTER dataset. When incorporating an additional Slovenian corpus into the training set, the multilingual model exhibits superior performance compared to state-of-the-art approaches in specific scenarios. Moreover, the newly introduced NOBI labeling mechanism enhances the classifier’s capacity to extract short nested terms significantly, leading to substantial improvements in Recall for the ACTER dataset and consequentially boosting the overall F1-score performance.

Isolating Terminology Layers in Complex Linguistic Environments: A Study about Waste Management

Automatic term extraction aims at extracting terminological units from specialized corpora to assist terminographers in developing glossaries, thesauri, and termbases. Unfortunately, traditional methods often overlook the complex relation between terminologies of different subject fields that co-occur in a single specialized corpus. This study illustrates Domain Concept Relatedness, a novel term extraction technique meant to isolate the terminology of a given subject field. We test our technique against the term extraction tool of Sketch Engine and the contrastive approach by applying them to the extraction of waste management terms from a new Italian corpus about waste management legislation. The results show that Domain Concept Relatedness effectively extracts multi-word terms belonging to a given subject field but still fails to extract single-word terms.

CHV.br: Exploratory study for the development of a consumer health vocabulary (CHV) supported by a network model for Brazilian Portuguese language

Successful consumer health vocabulary (CHV) models have been engineered and updated by using automatic term extraction techniques from online content. However, the relationship between terms has yet to be mapped. This study aims to describe a CHV model for the Brazilian Portuguese language that is supported by a complex network. The method was split up into three distinct stages: (1) collect and automatically extract terms from structured data sources on the web, such as Unified Medical Language System (UMLS) vocabularies and DBpedia; (2) construct a complex network; and (3) select the terms supported by clustering techniques. A model called CHV.br was developed and supported by a complex network structure which makes connections between the controlled vocabulary and consumer vocabulary and maps semantic relationships as categories, synonyms and related terms. CHV.br contains 146,956 terms, of which 31,439 are UMLS preferred terms and 83,279 are synonyms. The CHV.br is available and powered by Simple Knowledge Organization System and Resource Description Framework standards. The method used in this study showed to be valid for the selection of the candidate terms by connecting the terms from different reliable resources, in addition to expanding the number of terms and their semantic relationships. The content and structure of CHV.br could play a vital role in enhancing the development of consumer-oriented health applications.

A systematic review of Automatic Term Extraction: What happened in 2022?

Abstract Automatic Term Extraction (ATE) systems have been studied for many decades as, among other things, one of the most important tools for tasks such as information retrieval, sentiment analysis, named entity recognition, and others. The interest in this topic has even increased in recent years given the support and improvement of the new neural approaches. In this article, we present a follow-up on the discussions about the pipeline that allows extracting key terms from medical reports, presented at MDTT 2022, and analyze the very last papers about ATE in a systematic review fashion. We analyzed the journal and conference papers published in 2022 (and partially in 2023) about ATE and cluster them into subtopics according to the focus of the papers for a better presentation.

An unsupervised linguistic-based model for automatic glossary term extraction from a single PDF textbook

Term extraction from textbooks is the cornerstone of many different intelligent natural language processing systems, especially those that support learners and educators in the education system. This paper proposes a novel unsupervised domain-independent model that automatically extracts relevant and domain-related key terms from a single PDF textbook, without relying on a statistical technique or external knowledge base. It only relies on the basic linguistic techniques of the natural language processing: pattern recognition, sentence tokenization, part-of-speech tagging, and chunking. The model takes a PDF textbook as an input and produces a list of key terms as an output. Furthermore, the model proposes a novel classification of sentences from which the concept of defining sentences is proposed. The defining sentences are the main textual units that the model revolves around to identify the key terms. The architecture of the proposed work consists of 21 processes distributed across three phases. The first phase consists of five processes for extracting text from a PDF textbook and cleaning it for the next phases. The second phase consists of eight processes for identifying the defining sentences and extracting them from all the textbook’s sentences. The last phase consists of eight processes for identifying and extracting the key terms from every defining sentence. The proposed work was evaluated by two experiments in which two PDF textbooks from different fields are used. The experimental evaluation showed that the results were promising.

TV series as disseminators of emerging vocabulary: Non-codified expressions in the TV Corpus

Abstract This study presents a method for identifying words that appear in corpus data earlier than their first date of attestation in dictionaries. We demonstrate the application of this method based on a large diachronic corpus, the TV Corpus, and the Oxford English Dictionary (OED). Combining automatic extraction of candidate terms from the TV Corpus with comprehensive manual analysis and verification, the method identifies 32 words that were used in TV series before their first attestation in the OED. We present a detailed discussion of these words, analysing their distribution across decades and genres of the TV Corpus, their origins, semantic domains and word-formation processes. We also present extracts with their first uses in the TV Corpus and analyse how the words were presented to the large and anonymous mass audience. Our study shows that the method we present is suitable for identifying early attestations of words in large corpora, even though in the case of the TV Corpus, a great deal of manual analysis and verification is needed. In addition, we argue that TV series and other types of fictional texts are an important resource for studying the coinage and spread of terms, due to their function and the fact that they address a mass audience.

Automatic extraction of significant terms from the title and abstract of scientific papers using the machine learning algorithm: A multiple module approach

Keyword extraction is the task of identifying important terms or phrase that are most representative of the sourcedocument. Although the process of automatic extraction of keywords from title is an old method, it was mainly forextraction from a single web document. Our approach differs from previous research works on keyword extraction in severalaspects. For those who are non-expert of the scientific fields, understating scientific research trends is difficult. The purposeof this study is to develop an automatic method of obtaining overviews of a scientific field for non-experts by capturingresearch trends. This empirical study excavates significant term extraction using Natural Language Processing (NLP) tools.More than 15000 titles saved in a .csv file was our dataset and scripts written in Python were our process to compare how farsignificant terms of scientific title corpus are similar or different to the terms available in the abstract of that same scientificarticle corpus. A light-weight unsupervised title extractor, Yet Another Keyword Extractor (YAKE) was used to extract theresults. Based on our analysis, it can be concluded that these algorithms can be used for other fields too by the non-expertsof that subject field to perform automatic extraction of significant words and understanding trends. Our algorithm could be asolution to reduce the labour-intensive manual indexing process.

Chinese technical terminology extraction based on DC-value and information entropy

China's technology is developing rapidly, and the number of patent applications has surged. Therefore, there is an urgent need for technical managers and researchers that how to apply computer technology to conduct in-depth mining and analysis of lots of Chinese patent documents to efficiently use patent information, perform technological innovation and avoid R&D risks. Automatic term extraction is the basis of patent mining and analysis, but many existing approaches focus on extracting domain terms in English, which are difficult to extend to Chinese due to the distinctions between Chinese and English languages. At the same time, some common Chinese technical terminology extraction methods focus on the high-frequency characteristics, while technical domain correlation characteristic and the unithood feature of terminology are given less attention. Aiming at these problems, this paper proposes a Chinese technical terminology method based on DC-value and information entropy to achieve automatic extraction of technical terminology in Chinese patents. The empirical results show that the presented algorithm can effectively extract the technical terminology in Chinese patent literatures and has a better performance than the C-value method, the log-likelihood ratio method and the mutual information method, which has theoretical significance and practical application value.

TermitUp: Generation and enrichment of linked terminologies

Domain-specific terminologies play a central role in many language technology solutions. Substantial manual effort is still involved in the creation of such resources, and many of them are published in proprietary formats that cannot be easily reused in other applications. Automatic term extraction tools help alleviate this cumbersome task. However, their results are usually in the form of plain lists of terms or as unstructured data with limited linguistic information. Initiatives such as the Linguistic Linked Open Data cloud (LLOD) foster the publication of language resources in open structured formats, specifically RDF, and their linking to other resources on the Web of Data. In order to leverage the wealth of linguistic data in the LLOD and speed up the creation of linked terminological resources, we propose TermitUp, a service that generates enriched domain specific terminologies directly from corpora, and publishes them in open and structured formats. TermitUp is composed of five modules performing terminology extraction, terminology post-processing, terminology enrichment, term relation validation and RDF publication. As part of the pipeline implemented by this service, existing resources in the LLOD are linked with the resulting terminologies, contributing in this way to the population of the LLOD cloud. TermitUp has been used in the framework of European projects tackling different fields, such as the legal domain, with promising results. Different alternatives on how to model enriched terminologies are considered and good practices illustrated with examples are proposed.

Automatic medical term extraction from Vietnamese clinical texts

Abstract In this paper, we propose the first method for automatic Vietnamese medical term discovery and extraction from clinical texts. The method combines linguistic filtering based on our defined open patterns with nested term extraction and statistical ranking using C-value. It does not require annotated corpora, external data resources, parameter settings, or term length restriction. Beside its specialty in handling Vietnamese medical terms, another novelty is that it uses Pointwise Mutual Information to split nested terms and the disjunctive acceptance condition to extract them. Evaluated on real Vietnamese electronic medical records, it achieves a precision of about 74% and recall of about 92% and is proved stably effective with small datasets. It outperforms the previous works in the same category of not using annotated corpora and external data resources. Our method and empirical evaluation analysis can lay a foundation for further research and development in Vietnamese medical term discovery and extraction.

NMF-based approach to automatic term extraction

This work describes automatic term extraction approach based on the combination of the probabilistic topic modelling (PTM) and non-negative matrix factorization (NMF). Topic modeling algorithms including NMF-based ones do not require expensive and time-consuming manual annotations for domain terms, but only a corpus of domain documents. The topics emerge from the corpus documents without any supervision as sets of most probable words. This work is aimed to investigate how fully and precisely these most probable words from topics can reflect domain terminology. We run a series of experiments on the novel, qualitatively annotated dataset ACTER that was first used in the TermEval 2020 Shared Task. We compare five different NMF algorithms and four different NMF initializations when changing the number of topics extracted from documents and the number of most probable words extracted from topics in order to determine optimal combinations for best performance of term extraction. Finally, we compare the obtained optimal combinations of NMF with the competitive methods in TermEval 2020 and prove that our approach is second only to two much more sophisticated, domain-dependent supervised methods.

Automatic labeling and extraction of terms in natural language processing in acupuncture clinical literature

The paper analyzes the specificity of term recognition in acupuncture clinical literature and compares the advantages and disadvantages of three named entity recognition (NER) methods adopted in the field of traditional Chinese medicine. It is believed that the bi-directional long short-term memory networks-conditional random fields (Bi LSTM-CRF) may communicate the context information and complete NER by using less feature rules. This model is suitable for term recognition in acupuncture clinical literature. Based on this model, it is proposed that the process of term recognition in acupuncture clinical literature should include 4 aspects, i.e. literature pretreatment, sequence labeling, model training and effect evaluation, which provides an approach to the terminological structurization in acupuncture clinical literature.

Tagging terms in text

Abstract As with many tasks in natural language processing, automatic term extraction (ATE) is increasingly approached as a machine learning problem. So far, most machine learning approaches to ATE broadly follow the traditional hybrid methodology, by first extracting a list of unique candidate terms, and classifying these candidates based on the predicted probability that they are valid terms. However, with the rise of neural networks and word embeddings, the next development in ATE might be towards sequential approaches, i.e., classifying each occurrence of each token within its original context. To test the validity of such approaches for ATE, two sequential methodologies were developed, evaluated, and compared: one feature-based conditional random fields classifier and one embedding-based recurrent neural network. An additional comparison was added with a machine learning interpretation of the traditional approach. All systems were trained and evaluated on identical data in multiple languages and domains to identify their respective strengths and weaknesses. The sequential methodologies were proven to be valid approaches to ATE, and the neural network even outperformed the more traditional approach. Interestingly, a combination of multiple approaches can outperform all of them separately, showing new ways to push the state-of-the-art in ATE.

Особенности разметки научно-технических текстов в аспекте создания специализированного корпуса

The article deals with the peculiarities of the markup of scientific and technical texts in developing a corpus of highly specialized texts. The scientific and technical texts as sources of filling the corpus are listed. The scientific and technical texts are analyzed from the position of markup of textual elements of different levels. The necessity of introducing interlevel types of markup of scientific and technical texts is substantiated. The significance of introducing structural markup when creating a corpus of scientific and technical texts is emphasized. The structural elements of scientific and technical texts for filling the corpus are listed. The current state of the problem of automatic extraction of terms from scientific and technical texts is analyzed. It is shown that the greatest difficulty is the marking of multicomponent terminological units in the corpus of scientific and technical texts. We identify literary terms as objects that require the development of additional tools for their processing, which may include various letters, symbols, numbers or their combinations. References as a factor influencing the classification and rubrication of scientific and technical texts are analyzed. The necessity of studying the types of references, as well as the ways of their automatic marking in the corpus of scientific and technical texts is substantiated. The necessity of introducing a separate marking of examples in scientific and technical texts is substantiated.

Methodological Framework for the Development of an English-Lithuanian Cybersecurity Termbase

The aim of the paper is to present a methodological framework for the development of an English-Lithuanian bilingual termbase in the cybersecurity domain, which can be applied as a model for other language pairs and other specialised domains. It is argued that the presented methodological approach can ensure creation of high-quality bilingual termbases even with limited available resources. The paper touches upon the methods and problems of dataset (corpora) compilation, terminology annotation, automatic bilingual term extraction (BiTE) and alignment, knowledge-rich context extraction, and linguistic linked open data (LLOD) technologies. The paper presents theoretical considerations as well as the arguments on the effectiveness of the described methods. The theoretical analysis and a pilot study allow arguing that: 1) a combination of parallel and comparable corpora enable to considerably expand the amount and variety of data sources that can be used for terminology extraction; this methodology is especially important for less-resourced languages which often lack parallel data; 2) deep learning systems trained by using manually annotated data (gold standard corpora) allow effective automatization of extraction of terminological data and metadata, which enables to regularly update termbases with minimised manual input; 3) LLOD technologies enable to integrate the terminological data into the global linguistic data ecosystem and make it reusable, searchable and discoverable across the Web.

Method for Automatic Term Extraction from Scientific Articles Based on Weak Supervision

We propose a method for scientific terms extraction from the texts in Russian based on weakly supervised learning. This approach doesn't require a large amount of hand-labeled data. To implement this method we collected a list of terms in a semi-automatic way and then annotated texts of scientific articles with these terms. These texts we used to train a model. Then we used predictions of this model on another part of the text collection to extend the train set. The second model was trained on both text collections: annotated with a dictionary and by a second model. Obtained results showed that giving additional data, annotated even in an automatic way, improves the quality of scientific terms extraction.

Multilingual Automatic Term Extraction in Low-Resource Domains

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.

Building knowledge graphs of regulatory documentation based on semantic modeling and automatic term extraction

The paper proposes a new complex solution for automatic analysis and terms identification in regulatory and technical documentation (RTD). The task of terms identification in the documentation is one of the key issues in the digitalization dealing with the design and construction of buildings and structures. At the moment, the search and verification of RTD requirements is performed manually, which entails a significant number of errors. Automation of such tasks will significantly improve the quality of computer-aided design. The developed algorithm is based on such methods of natural language analysis as tokenization, search for lemmas and stems, analysis of stop words and word embeddings applied to tokens and phrases, part-of-speech tagging, syntactic annotation, etc. The experiments on the automatic extraction of terms from regulatory documents have shown great prospects of the proposed algorithm and its application for building knowledge graphs in the design domain. The recognition accuracy for 202 documents selected by experts was 79 % for the coincidence of names and 37 % for the coincidence of term identifiers. This is a comparable result with the known approaches to solving this problem. The results of the work can be used in computer-aided design systems based on Building information modeling (BIM) models, as well as to automate the examination of design documentation.

Изучение терминологических подсистем современных школьных учебников на русском языке с помощью модели анализа семантики естественных языков Word2Vec

The article reports the results of the study that explored the inventory and functioning of scientific terms and special lexemes in textbooks for Russia’s secondary schools. The toolset included modern methods of natural language processing and deep learning. The number of terms from different fields of knowledge that a secondary school student should learn has never been evaluated. According to the preliminary evaluations based on the Model Basic Curriculum for General and Secondary Education 2015, a secondary school leaver is supposed to be able to understand, recognise and use about 1,000 terms and terminological combinations in the subject Russian Language alone. Thus, taking into account the number of school subjects, the total number of special vocabulary studied in general education schools is measured in thousands. At the same time, the comparative characteristics of the inventory and functioning of terms in textbooks for different school subjects are under-scrutinized and remain unknown. Besides, it is unclear how the terminological density of school textbooks for different subjects correlates with the place occupied by these subjects in the curriculum. The traditional way of compiling lists of special terms is simply to glean them from special texts and write down manually. This method is reliable to gain insights into the best selection practices, however, it cannot be applied to large data sets and does not reflect the term frequency, the specificity of their syntagmatic connections, or the systemic relationship between them. Our project is aimed at filling this gap through: 1) creating a full-text corpus of school textbooks approved by the Ministry of Education for grades 5–11, 2) automatic extraction, stratification, and mapping of terms with the help of distribution semantics algorithms, 3) creation and training of a deep neural network capable of predicting the subject, level of education and educational topic given a group of vector theoretical development of terminology science. They may also find practical application, e. g., in the development of different types of educational literature.