Deep Learning Methods for Identification of Multiword Flower and Plant Names

Proceedings of Recent Advances in Natural Language Processing, pages 879–887 Varna, Sep 4–6, 2023 https://doi.org/10.26615/978-954-452-092-2_095 879 Deep Learning Methods for Identification of Multiword Flower and Plant Names Damith Premasiri1, Amal Haddad Haddad2, Tharindu Ranasinghe3 and Ruslan Mitkov4 1University of Wolverhampton, UK 2University of Granada, Spain 3Aston University, UK 4Lancaster University, UK damith.premasiri@wlv.ac.uk, amalhaddad@ugr.es t.ranasinghe@aston.ac.uk, r.mitkov@lancaster.ac.uk Abstract Multiword Terms (MWTs) are domain-specific Multiword Expressions (MWE) (Paji´c et al., 2018) where two or more lexemes converge to form a new unit of meaning (León Araúz and Cabezas García, 2020). The task of processing MWTs is crucial in many Natural Language Processing (NLP) applications, including Ma- chine Translation (MT) and terminology extrac- tion. However, the automatic detection of those terms is a difficult task and more research is still required to give more insightful and useful results in this field. In this study, we seek to fill this gap by using state-of-the-art transformer models. We evaluate both BERT (Devlin et al., 2019) like discriminative transformer models and generative pre-trained transformer (GPT) (Radford et al., 2018) models on this task, and we show that discriminative models perform better than current GPT models in the identifi- cation of multiword flower and plant names for both English and Spanish. Best discriminative models perform with 94.3, 82.1 F1 scores in English and Spanish data, respectively, while ChatGPT could only return 63.3 and 47.7 F1 scores, respectively. 1 Introduction Botany is a multidisciplinary field that encom- passes different scientific disciplines such as Ge- netics, Ecology, Physiology, Biochemistry, Archi- tecture, Gastronomy, Commerce, Art and Design, etc. One of the key areas in Botany is the study of flowers and plants. The market of flowers and plants is regarded as an economic engine of dif- ferent economic and industrial activities. For this reason, its study and analysis are considered rele- vant in order to make this domain more accessible to all users, both at scientific and professional lev- els and also at layperson level, as flowers and plants have important national and international symbol- isms and their roots are profoundly embedded in cultures and societies. The accurate identification and denomination of each plant is essential for the correct development and dissemination of science in all those multidisciplinary fields. It is also cru- cial for the correct communication of knowledge in different languages and also for the proper design of lexicographic resources and thesaurus. From the point of view of applied linguistics, the identification of names of flowers and plants is relevant to language professionals. From a ter- minological point of view, it helps in laying the basis of term coining processes and gives insights into the underlying mechanisms of term creation. Translators also benefit from this information for the translation process. Taking into consideration the quick develop- ment of NLP technologies and the importance of Machine translation (MT) in the dissemination of knowledge and in building new resources, it is im- portant to extend the studies and cover new areas of research, such as Botany. The automatic iden- tification of …

Content truncated to 500 words