Опыт использования ресурса Twee при работе с новостными статьями на занятиях по профессионально ориентированному английскому языку

Полный текст

Problem statement. Any professional needs to keep up to date with what is happening in their field of activity both domestically and internationally. In this regard, it is especially important for future graduates to be able to work with various information resources, in which text materials in the original language from various periodicals occupy a significant place. It is natural that foreign language teachers in higher education, trying to meet the above-mentioned needs of students, often organize classes with the use of various news resources, including articles, which are processed and supported with different language exercises. However, sourcing and preparing for a class using news articles often takes a lot of time [1], not to mention the fact that news materials themselves need to be updated on a regular basis. In this regard, the use of a number of modern tools based on AI can greatly assist in arranging classes and facilitate the work of teachers including those of Maritime English. These are such tools as, for example, Quillbot, Diffit and Twee, the experience of using the last of which will be covered in this paper. News articles have long been used by language teachers as learning materials in the classroom. As early as the 20th century, researchers worldwide acknowledged the significance of incorporating authentic periodicals into the teaching process [2-4]. Most researchers have always emphasized that the application of news content in higher education classes is necessary for future professionals to have information about what is happening in the world in their respective field regardless of its nature, ranging from IT [5] to journalism and media studies [6; 7], from climate studies [8] to medicine [9]. At the same time, the usage of foreign news resources also seems important from the linguistic point of view, helping students to expand and consolidate their vocabulary, master lexical and grammatical material [10; 11]. Moreover, a number of researchers base their work on analyses of particular news outlets. Thus, J. Kapadiya explores the use of The Times of India newspaper to develop students’ speaking and writing skills [12], U. Shamin and S. Shoukat regard Dawn, a Pakistan-based news outlet, as a vocabulary enrichment tool [13], while E. Mohamed, investigating the issue of introducing new terms into students’ vocabulary, takes the world-famous British newspaper The Guardian as an example [14]. AI-based platforms, such as Twee or Diffit, are now extensively used by foreign language teachers to facilitate preparation for classes, including reading-oriented ones, as M. Boeru and other researchers report in their works [15]. However, although the use of AI-based platforms is widely discussed within English education, there are still not many detailed research results on their usage within the context of vocational, namely Maritime English [16; 17]. Moreover, the use of AI resources to deal with maritime news articles in English classes is currently left almost unexplored. Methodology. At present, there are a number of foreign and domestic professional-oriented news resources that are published in electronic form and do not require either user fees or mandatory registration. Speaking about the maritime sphere, among quite a large number of news magazines the following can be distinguished: British magazines Shipping Today and Lloyds’s List (the second being one of the oldest continuously running specialized magazines), Shipping Gazette, Sea News (a Russian specialized magazine covering mainly domestic news in the maritime sphere with material duplicated in English) as well as the American resource The Maritime Executive, which will be described in more detail. Thus, The Maritime Executive has proved to be a useful resource for undergraduate Transportation Process Technology students. It updates news information on a regular basis, six days a week, with four top stories in each issue, includes a features genre, and, importantly, has an article rubricator that allows the teacher to select an article of a particular thematic category. For example, depending on the topic being studied, it is possible to select relevant articles from such categories as ‘Shipping’ and ‘Ports’, which are probably the most important for future logisticians, or, for example, from ‘Shipbuilding’, ‘Environment’, ‘Cruise ships’, ‘Salvage’, and others. It is also important to note that the articles on this resource are accompanied with audio tracks, so that the teacher can create a number of listening activities on this basis, being able to adjust the speed of text reproduction. For example, it is possible to print out an article, removing certain vocabulary currently being studied, or, alternatively, excluding numerals from the text so that students fill in the missing linguistic units by ear, etc. However, it should be remembered that in addition to lexical and grammatical tasks, news articles can be used to make up a number of speaking and writing activities, and in this regard, the content of the articles should be taken into account. If one of the tasks involves retelling, texts full of abbreviations, proper names and numerals might not be quite suitable for this purpose. For successful development of coherent speech skills, it is more appropriate to choose articles with event content such as news reporting on incidents at sea, constructing and launching new vessels, highlighting environmental problems, etc. Having listed the main maritime news outlets, let us consider the Twee platform as one of the resources for processing news articles with emphasis on the structure of the class considering that assignments based on news articles like any other texts should follow a certain sequence [18], regardless of whether AI tools are used or they are created manually. In the first stage, students are given a pre-reading task, the purpose of which is to motivate them to read the material and to possibly identify their existing knowledge on the topic [19]. The pre-reading task is followed by reading itself, which is useful to combine with simultaneous listening to the text [20] or with the listening gap-fill assignment mentioned above. This is continued by readingcomprehension tasks [21] and vocabulary and grammar tasks which are usually several in number. At the final stage, students are encouraged to complete follow-up tasks aimed at developing coherent spoken and written language and, in this case, the written tasks may well be classed as extracurricular activities [22]. The experience of creating tasks using the Twee resource for senior undergraduate students studying in the field of ‘Transport Process Technology’ has shown that this tool successfully copes with generating exercises based on news texts. The tasks for the pre-reading stage and follow-up assignments seem to be particularly successful. For example, before reading an article describing a ship capsize allegedly caused by improper loading of bulk cargo[13], the Twee platform suggests showing students pictures of different vessel types and initiating a short discussion on ‘advantages and disadvantages of a vessel, depending on its type’, with particular attention to the issue of safety. Before reading an article on the environmental problem of marine and animal pollution by oil products[14], the platform suggests showing students a photo of an oil-contaminated bird and eliciting their verbalised reactions. As part of the pre-reading tasks for an article covering the Baltimore bridge collapse tragedy[15], Twee suggests drawing on students’ personal experiences and asking them a series of questions, one of which is whether they have ever witnessed a transport accident. Notably, the resource usually offers several options for pre-reading tasks, and it is left to the teacher to choose the one they consider to be the most appropriate. The advantage of this type of task offered on the Twee platform, apart from their various alternatives, is that they do not require any additional effort on the part of users, except for the selection of appropriate illustrative material. As follow-up tasks that can often be provided for students to complete outside the classroom, the Twee platform offers a number of written assignments that include writing an essay, a business or personal letter, a blog post, a review, or an advertisement on a given topic, among others. The above assignments are accompanied with a list of required vocabulary, which, according to the task description, should be used throughout the written work. A useful aid for the teacher is the possibility to choose the level of students’ language proficiency from A1 to C2 [23]. Depending on the selected level, the complexity of the wording of both the tasks and the phrases created by the system to practise the lexical and grammatical material will vary. For example, when creating a ‘Matching the definitions’ task, the ‘crew’ term is defined for A1-A2 as ‘a group of people who work together on a ship’, as ‘a group of people who work together on a ship, airplane, or other vehicle’ for B1-B2, and as ‘a group of individuals who work together, especially on a ship or aircraft, performing specific roles’ for those at C1-C2. Alternatively, as pre-reading tasks for an article about a bridge collapse in Baltimore, the Twee platform offers the following questions for A1-A2, B1-B2 and C1-C2 level, respectively: 1. Have you ever seen a ship collision or another transportation accident? 2. Have you ever experienced a situation where safety measures failed? What happened? 3. Have you ever experienced a situation where a lack of proper maintenance led to a significant failure? The above questions demonstrate the gradual change in the level of text complexity, the entry of new lexical units as well as the transformation of the grammatical structure of a phrase. However, it should be understood that the language level division can often be only nominal due to the peculiarities of the uploaded material, since it is obvious that the system will hardly be able to create formulations completely corresponding to the A1 level for a non-adapted news article full of specialised vocabulary and respective verbal patterns even though the simplification of lexical and grammatical content is definitely observed. Some of the tasks listed above can be generated by the platform completely automatically. These are the aforementioned pre-reading and reading comprehension tasks. While pre-reading assignments can always be chosen by the teacher from those offered as most appropriate and are definitely a great support for the teacher, reading comprehension assignments may require some improvement. For reading comprehension activities, the Twee system offers standard ‘true/false’, ‘multiple choice’ and ‘open questions’ tasks. The answers to these are often quite obvious, and in the case of ‘open questions’ task, several questions may repeat each other in meaning, even if worded differently. However, as practice has shown, the platform is the least successful in handling multiple choice tasks. As an example, below is a question of this type and the answers to it generated by the Twee system for an article on a ship colliding with a bridge support in Baltimore[16]: What was the estimated gross tonnage of the container ship Dali? a 2.6 km b 300 m c 50 m d 95.000 tons Even with no knowledge of specialised vocabulary or event-based material, students are likely to give the correct answer to this question. To do so, it is enough to have the understanding that the tonnage of a ship cannot be expressed in terms of length. However, despite the evidence of the correct answer, even this sort of question may have some relevance, especially for students with poor language skills, providing them with an opportunity to choose the right one and, thus, creating a positive motivation for further material mastering. As for grammar and vocabulary tasks, the Twee system usually creates them in the following way. First the teacher enters the pre-specified lexical units, which are then used by the platform to create separate phrases. Next, there are a number of options, some of which are as follows: it is possible to manually remove these units from the created text so that the students could fill in the gaps with the required vocabulary; or else the platform can be asked to mix the words in the resulting phrases and thus create a ‘reconstruct the word order’ task. Finally, grammatical forms could be removed from the phrases created by the platform for students to complete the resulting ‘open the brackets’ task. In addition to selecting appropriate lexical units, the teacher should also identify polysemous vocabulary and indicate its required meaning in the input field to generate exercises. The issue is that the Twee system does not always trace the connection between the input words and the lexis of the news stories. For example, in one of the articles about the search for containers lost at sea5, the noun ‘Scavenger’ is used in the sense of ‘a person who searches for discarded items’, but the Twee platform gives the first meaning of this lexical unit when creating the ‘Match the definitions’ exercise, namely ‘an animal that feeds on dead animals or other decaying matter’. Finally, it is up to the teacher to decide which types of assignments are preferable for a particular news article and in what sequence they should be offered to learners. Results and discussion. Tables 1 and 2 below show the usage frequency of Twee and similar AI-based resources among English language teachers at nonlanguage universities and its evaluation. A total of 50 teachers were surveyed, most of whom are representatives of the Department of English for Navigation and Communication at the State University of Maritime and Inland Shipping, St. Petersburg, Russia. Table 1 Usage frequency of artificial intelligence-based platforms among teachers at non-linguistic universities The frequency and nature of AI-based platform usage Number of respondents Use Twee or other learning platforms on a regular basis 35 Have short-term experience with AI-based platforms 7 Would like to start using AI-based platforms 6 Do not use either Twee or any other platforms and have no intention of doing so 2 Source: compiled by Varvara S. Golubeva. 5 Salvage Firm Fined for Scavenging WWI Shipwreck. The Maritime Executive. Available from: https://www.maritime-executive.com/article/salvage-firm-fined-for-scavenging-wwi-shipwreck (accessed: 06.11.2025). Table 2 Evaluation of AI-based platforms usage by non-language university teachers Evaluating the use of AI-based platforms Number of respondents Are completely satisfied 22 Are generally satisfied except for a limited number of issues 11 Are only partially satisfied, and resort to other ways of assignment creation 9 Are totally dissatisfied 0 Source: compiled by Varvara S. Golubeva. As can be seen from the results obtained, AI-based platforms, in particular Twee, are currently quite popular among English language teachers in higher education with the exception of only a small percentage of respondents not aiming to use any AI platforms, and largely fulfil their needs, although there are a number of aspects that, from the questioned respondents’ point of view, require some improvement. Both the positive and problematic issues regarding the use of the Twee resource for generating assignments based on specialized texts summarised in the Table 3 are based on the previously described experience of implementing this resource for class preparation. Table 3 Positive aspects and problem areas of the Twee platform in terms of text-based task generation Task type Positive aspects Aspects to be improved Pre-reading A wide range of options is available. No teacher involvement is required, except for selecting illustrative material, if necessary. Quick task generation. Tasks are adaptable to the students’ language proficiency level Reading comprehension No teacher involvement is required. Quick task generation. Tasks are adaptable to the students’ language proficiency level The answers to questions may be obvious, especially in case of ‘multiple choice’ task Grammar and vocabulary Quick task generation. Tasks are adaptable to the students’ language proficiency level Does not always provide the required meaning of a lexical unit. Requires the teacher to select key vocabulary from the text Follow-up Quick task generation. A wide range of options is available. Only minimal teacher involvement is required. Tasks are adaptable to the students’ language proficiency level Source: compiled by Varvara S. Golubeva. It is worth noting that, despite the undeniable benefits that Twee and similar resources bring to the process of teaching a foreign language and to preparing for classes, it is unlikely that the entire process should be completely outsourced to Twee. Therefore, it is not surprising that there are still debates and surveys being conducted in the academic community as to whether AI-powered tasks are actually superior to those created by humans [24]. There is no reason to assume that an AI-based platform is able (at least for the time being) to fully automate the process of creating exercises for the given material so that teachers could be completely relieved of class preparation. And while the platform with varying results is able to create reading comprehension tasks, as well as pre-reading and follow-up assignments without human assistance, this is not the case with lexical and grammar exercises. For example, teachers need to make their own decision about which lexical units from the text to include in the exercises and then prompt the system to create certain tasks on that basis. Also, it would be valuable practice for teachers to occasionally create class assignments based on news articles or other materials manually in order to retain this skill. Such a manual process is more likely to preserve the ability to think critically and to analyse, as well as to ensure that the quality of teaching is maintained should AI-based resources become unavailable. This recommendation seems particularly relevant for new professionals in teaching who may not yet have had the opportunity to polish their skills in creating their own assignments in class preparation. Summarising the above, AI functions should be used only after mastering the skill of handcrafting exercises. In addition, it should be added that manual exercise creation, although often labour-intensive and time-consuming, can be an exciting activity that only requires some immersion. Conclusion. Based on the practical use of the Twee resource in Vocational English classes, it may be concluded that this tool is obviously a significant advantage for teachers in preparing for classes, reducing their time and effort. The Twee platform with minimal involvement of a teacher helps to create classes, including those based on news articles, generating tasks for each stage of the class such as pre-reading, reading comprehension, grammar and vocabulary and followup tasks. However, as the experience of using tasks created by Twee has shown, the results obtained from this resource should be thoroughly checked and analysed before offering them to the students. Particular attention should be paid to the adequacy of the assignments to the task set in the lesson, the correspondence of the tasks and their formulation to the students’ language level, as well as the correspondence of definitions to the given lexical units. In general, the success of using this resource in reading-focused classes largely depends on the teacher, since it is he/she who selects the appropriate news content, determines the type of tasks and the order to complete them, as well as chooses the necessary lexical units to include in the assignments.

Об авторах

Варвара Сергеевна Голубева

Государственный университет морского и речного флота им. адмирала С.О. Макарова

Автор, ответственный за переписку.
Email: golubevavs@gumrf.ru
ORCID iD: 0000-0001-9320-1730
SPIN-код: 5959-5730

старший преподаватель кафедры английского языка навигации и связи

Российская Федерация, 198035, Санкт-Петербург, ул. Двинская, д. 5/7

Список литературы

Дополнительные файлы