Network-Enabled Keyword Extraction for Under-Resourced Languages

Beliga, Slobodan; Martinčić-Ipšić, Sanda

doi:10.1007/978-3-319-53640-8_11

Cited by 5 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…6, г. Тюмень, 625003, Россия. 2 Новосибирский национальный исследовательский государственный университет, ул. Пирогова, д.…”

Section: Artificial Intelligenceunclassified

“…В настоящее время большинство методов подбора ключевых слов протестировано на англоязычных текстовых корпусах, в то время как для анализа русскоязычных текстов используется достаточно узкий набор методов выделения ключевых слов [1]. Существует несколько подходов к подбору ключевых слов: извлечение ключевых слов непосредственно из текста, подбор ключевых слов из заранее определенного перечня тематик или рубрик и генерация ключевых слов на основе семантики текста путем его обобщения и перефразирования [2,3]. В последнем случае задача подбора ключевых слов схожа с задачей автоматического абстрактного реферирования текстов.…”

Section: Introductionunclassified

See 1 more Smart Citation

Keyphrase generation for the Russian-language scientific texts using mT5

Glazkova,

Morozov,

Vorobeva

et al. 2023

Model. anal. inf. sist.

View full text Add to dashboard Cite

In this work, we applied the multilingual text-to-text transformer (mT5) to the task of keyphrase generation for Russian scientific texts using the Keyphrases CS&Math Russian corpus. The automatic selection of keyphrases is a relevant task of natural language processing since keyphrases help readers find the article easily and facilitate the systematization of scientific texts. In this paper, the task of keyphrase selection is considered as a text summarization task. The mT5 model was fine-tuned on the texts of abstracts of Russian research papers. We used abstracts as an input of the model and lists of keyphrases separated with commas as an output. The results of mT5 were compared with several baselines, including TopicRank, YAKE!, RuTermExtract, and KeyBERT. The results are reported in terms of the full-match F1-score, ROUGE-1, and BERTScore. The best results on the test set were obtained by mT5 and RuTermExtract. The highest F1-score is demonstrated by mT5 (11,24 %), exceeding RuTermExtract by 0,22 %. RuTermextract shows the highest score for ROUGE-1 (15,12 %). According to BERTScore, the best results were also obtained using these methods: mT5 — 76,89 % (BERTScore using mBERT), RuTermExtract — 75,8 % (BERTScore using ruSciBERT). Moreover, we evaluated the capability of mT5 for predicting the keyphrases that are absent in the source text. The important limitations of the proposed approach are the necessity of having a training sample for fine-tuning and probably limited suitability of the fine-tuned model in cross-domain settings. The advantages of keyphrase generation using pre-trained mT5 are the absence of the need for defining the number and length of keyphrases and normalizing produced keyphrases, which is important for flective languages, and the ability to generate keyphrases that are not presented in the text explicitly.

show abstract

Section: Artificial Intelligenceunclassified

Section: Introductionunclassified

Keyphrase generation for the Russian-language scientific texts using mT5

Glazkova,

Morozov,

Vorobeva

et al. 2023

Model. anal. inf. sist.

View full text Add to dashboard Cite

show abstract

“…Poznat je samo mali broj istraživanja koji je proveden za neke od jezika poput španjolskog i francuskog (O. Medelyan, 2009), portugalskog (Abilhoa, De Castro, 2014), češkog (Kozłowski, Kozłowski, 2014), hrvatskog (Beliga, Martinčić-Ipšić, 2017;Beliga, Meštrović, Martinčić-Ipšić, 2016), srpskog (Beliga, Kitanović, Stanković, Martinčić-Ipšić, 2018) i sl. Međutim, ta istraživanja nužno ne uspoređuju uspješnost dvije spomenute metode, već se pojedina od njih bave specifičnom problematikom istraživanja vezanog uz ekstrakciju ključnih riječi.…”

Section: Uvodunclassified

Automatska ekstrakcija ključnih riječi iz teksta standardnim računalnim postupcima

Beliga

2020

Politehnika (Online)

Self Cite

View full text Add to dashboard Cite

Automatska ekstrakcija ključnih riječi iz teksta aktualan je istraživački problem u području računalne analize prirodnog jezika i pretraživanja informacija. Iako su razvijene brojne metode za ekstrakciju ključnih riječi iz teksta, njihova učinkovitost ovisna je o brojnim faktorima poput pristupa kojim su konstruirane, domene na koju su prilagođene, vrste jezika ili zadataka za koji su konstruirane i sl., a samim time prostor za napredak u smislu nadogradnje i poboljšanja, uvijek postoji. U ovom radu objašnjene su i rekonstruirane dvije postojeće metode – RAKE i MAUI, a koje su standardni predstavnici nenadzirane i nadzirane skupine metoda. Eksperimentalno je ispitano mogu li metode uspješno ekstrahirati ključne riječi iz tekstova pisanih na talijanskom jeziku, na kojem do sada nisu usporedno testirane. Za potrebe eksperimenta prikupljeni su i ručno označeni talijanski tekstovi. Efikasnost MAUI metode pokazala se perspektivnijom u odnosu na RAKE metodu što je već ranije potvrđeno u eksperimentu ekstrakcije ključnih riječi iz tekstova pisanih na engleskom jeziku.

show abstract

“…This issue has been particularly problematic for academic search tasks (Hoeber and Khazaei, 2015; Magara et al , 2017) and is central to the current study. The ubiquity of this challenge is evidenced by the number of extraction and analysis methods to improve queries – some of which have been reviewed in Beliga (2014), Beliga et al (2015) and Gupta (2017). KE and KW are domain-independent; however, they have been applied extensively to identify trends in research-oriented publications.…”

Section: Rationalementioning

confidence: 99%

Digital library keyword analysis for visualization education research

Bratt

2018

JARHE

View full text Add to dashboard Cite

Purpose The purpose of this paper is to assess the efficacy of the Institute of Electrical and Electronics Engineers (IEEE) Xplore digital library search engine to return relevant materials on information visualization pedagogy literature and to recommend search strategies to assist the digital library academic readership improve the efficacy of their search tasks. Furthermore, the results are of interest to general readers using similar digital repositories. Design/methodology/approach An initial scoping review using EBSCO Discovery services returned the number and accessibility of sources and publications-based various Boolean searches. A revised search strategy focused the search to IEEE publications as the primary source of visualization research. A corpus of keywords were extracted from the 44 relevant articles and analyzed for relevance, keyword trends and contexts of use. Findings Keyword analysis results show visualization education research is confounded by several information retrieval issues: relevancy, incomplete taxonomy, non-standard lexicon, diverse disciplines and under-representation. Recommendations include: search strategies, alternative digital collections, a potential opportunity for research in information visualization pedagogy to address this gap in an emerging field and the need for more effective interactive tools to assist with keyword selection. Research limitations/implications The study focused on the IEEE publications as the primary source of visualization research. Practical implications A repository of visualization education research that is easily findable and relevant benefits both faculty using information visualization in their teaching and academics whose work must be disseminated to the broadest audience. Strategic keyword selection, interactive keyword tools or more robust thesaurus will enable IEEE Xplore digital library users to optimize their interaction with the system. Furthermore, results suggest a need for more research in information visualization pedagogy. Originality/value This is the only study to uniquely assess the efficacy of the IEEE Xplore digital library database system to retrieve relevant visualization education literature based on keyword search.

show abstract

Network-Enabled Keyword Extraction for Under-Resourced Languages

Cited by 5 publications

References 13 publications

Keyphrase generation for the Russian-language scientific texts using mT5

Keyphrase generation for the Russian-language scientific texts using mT5

Automatska ekstrakcija ključnih riječi iz teksta standardnim računalnim postupcima

Digital library keyword analysis for visualization education research

Contact Info

Product

Resources

About