A Review on Question Generation from Natural Language Text

Question generation is an important yet challenging problem in Artificial Intelligence (AI), which aims to generate natural and relevant questions from various input formats, e.g., natural language text, structure database, knowledge base, and image. In this article, we focus on question generation from natural language text, which has received tremendous interest in recent years due to the widespread applications such as data augmentation for question answering systems. During the past decades, many different question generation models have been proposed, from traditional rule-based methods to advanced neural network-based methods. Since there have been a large variety of research works proposed, we believe it is the right time to summarize the current status, learn from existing methodologies, and gain some insights for future development. In contrast to existing reviews, in this survey, we try to provide a more comprehensive taxonomy of question generation tasks from three different perspectives, i.e., the types of the input context text, the target answer, and the generated question. We take a deep look into existing models from different dimensions to analyze their underlying ideas, major design principles, and training strategies We compare these models through benchmark tasks to obtain an empirical understanding of the existing techniques. Moreover, we discuss what is missing in the current literature and what are the promising and desired future directions.

Print-based multimodal texts: Using illustrated poems for generating ideas and writing narratives

The article provides insights on how print-based multimodal texts can be utilized to generate ideas and help students to write narrative texts. A qualitative approach in research design was employed with a pre-test and a writing assignment. NAPLAN’s (2010) Writing-Narrative Marking Guide was adapted to evaluate the respondents’ narrative essays. Moreover, diary notes were used as instruments to gather data. The respondents were five Year 10 students from a suburban secondary school in Selangor, Malaysia. The findings indicated that the illustrated poem ‘Pond’ enabled the respondents to generate ideas for their writing. The narratives also fulfilled the criteria of text structure, ideas, character and setting, and vocabulary as stated in the Analytical Rubric for Narrative Marking. Therefore, the respondents were able to write narratives. However, the utilisation of picture stimulus only helped to generate ideas to a certain extent. The quality of narrative writing was also further improved through feedback provided during the teaching and learning sessions. The findings also revealed that the respondents were able to achieve the ‘can do’ statement as stated in the CEFR Assessment Format. Furthermore, the use of print-based multimodal texts encouraged multimodal and visual literacy as the respondents exploited semiotic resources within the texts. This study suggested that print-based multimodal texts could be used as instructional materials in narrative writing; thus, teachers should consider them to promote effective learning.

What textbooks offer and what teachers teach: an analysis of the Dutch reading comprehension curriculum

In the Netherlands, the quality of the reading curriculum is currently under debate because of disappointing results on national and international assessments of students’ reading skills and motivation. In a mixed-method study, we analyzed the content of Dutch textbooks for reading comprehension instruction (i.e., the implemented curriculum) and teachers’ evaluation and use of these books (i.e., the enacted curriculum). A materials analysis of reading comprehension lessons (N = 80) in eight textbooks for grades 4 and 5 was complemented with semi-structured teacher interviews (N = 29) and lesson observations (N = 11), with a focus on the quality of reading strategy and text structure instruction in the curriculum. Main findings are (1) a lack of alignment between lesson goals, theory, and assignments, (2) a strong focus on practicing strategies, (3) limited declarative knowledge about strategies and text structure, (4) little opportunities for self-regulated strategy application. The teachers that were interviewed mention similar problems, but still hardly deviate from the textbook’s content and pedagogical guidelines. We make recommendations to improve the quality of the curriculum.

Political Discourse as Sliding Mode Manifestations

The article considers the problem of political discourse transformation as exemplified by European (British&Russian) political rhetoric of the 20th century. Considered is the complex nature of political discourse comprising both its variable and invariable aspects reflecting discursive constancy and change which are regarded as manifestations of the sliding mode usually applicable to exact sciences phenomena. The major factors of transformation depend on social change caused by dramatic events in history, namely political conflicts. The invariable part of political discourse is concentrated around the text structure with dicteme as the main information and structure unit of the text and discourse. The variable part is determined by factors of speech regulation including target content of the utterance, status of the speaker and the listener, pre-supposition and post-supposition. Genre and register specificity of political discourse as its constant characteristics reflect the change. Conclusions offer some generalizations Virtual Learning offer for Biologic Informatics aspects

Conflict Mobilization in Russian Social Media Communities As an Actual Practice of Protest Communication: Collective and Personal Dimension

From the standpoint of communicative approach conflict mobilization is considered in the article as an actual practice of protest communication in modern Russia. The parameters of the collective-personal dimension of conflict mobilization in the most popular social networks among Russian users: VKontakte, Odnoklassniki, Instagram, Twitter, Facebook are presented. Major forms of social group members online participation in protest actions, interaction means of conflict mobilization communities and the formation of Internet solidarity have been characterized with such criteria as publication volume and frequency. Selective factual and emotional-psychological argumentation with elements of defamation of opponents is concluded to prevail in the communicative behavior of the addresser. Frequent implementation of phatic strategy through consolidating appeals to community members with elements of idealization and glorification of protest actors is revealed. The addresser's self-presentation strategy is noted to be applied rather rarely; it is actualized mainly through their identification with the collective position of the group and the mass addressee. It has been established that the addressee is characterized by active self-presentation through the presentation of a subjective position opposed to a mutual opponent, is distinguished by an appeal to ethical and emotional argumentation, a variety of phatic tactics. The article reveals the relevant language means of implementing the identified strategies, shows the features of the text structure of mobilization posts.

Kemampuan Menulis Teks Cerita pada Mahasiswa Program Sarjana Pendidikan Bahasa Indonesia Fakultas Keguruan dan Ilmu Pendidikan Universitas Bengkulu

The purpose of this study is to describe the ability to write story texts for students of the Indonesian Language Education Undergraduate Program FKIP Bengkulu University. The scope of this research covers aspects of (1) the suitability of the theme with the content, (2) the structure of the story text, (3) the choice of words (diction), (4) the use of story-style sentences, and (5) the Indonesian Language spelling and handwriting. This research uses descriptive method and quantitative approach. The research population was all students of the Indonesian Language Education Study Program FKIP Bengkulu University in the odd semester of 2018/2019 and the sample was purposive sampling, determining that there were 34 students in semester III-A. Collecting data uses test techniques and instruments in the form of instructions for writing stories with the theme of "love" and guidelines for assessing the ability to write stories based on five components, including content, text structure, diction, sentences, and mechanics. Data analysis used descriptive statistics to find the mean and determined qualitative results. The results showed that the ability to write story texts for the students of the Indonesian Language Education Undergraduate Program, FKIP Bengkulu University, was categorized as good. This is based on the score of all components with an average of 73.97. These results still require efforts to improve to be better. The results of each aspect of the ability to write story texts, namely aspects (1) the suitability of the theme with the content, good category (18.94), (2) story structure, good category (16.09), (3) accuracy of diction in good category (score 14 ,3), (4) narrative language style sentences are in good category (17.24), and (5) mechanics (BI spelling and handwriting) are in good category (7.32).

Teaching Design of Integrated Course of Reading and Writing Based on In-depth Learning in Senior High School — Take Unit 3 Computers as an Example

In-depth learning is a learning method, which can transfer knowledge and solve problems in various complex and real situations through learners' critical understanding and thinking towards learning content. Guided by the subject core competence, the integrated course of English reading and writing should combine autonomous and cooperative learning to promote the perceptual understanding, and then design multi angle writing training such as text structure imitation, plot rewriting and ending continuation, so as to promote the application and transfer of language knowledge to promote students' in-depth learning.

A CRITICAL DISCOURSE ANALYSIS OF A POLITICAL TALK SHOW ON THE 2019 INDONESIAN PRESIDENTIAL ELECTION

Political discourse on the 2019 Indonesian presidential election has been studied from a variety of linguistic viewpoints. However, the pros and cons discourse on Joko Widodo's coalition with Prabowo in spite of their rivalry has not been examined through critical discourse analysis. This study aims to explore the text structure, social cognition, and social context of the pros dan cons discourse as the theme of Mata Najwa talk show. This is a kind of qualitative analytical research using van Dijk’s (1996, 2015) critical discourse analysis framework. The data comprised utterances of the talk show participants examined through three levels of analysis: description of text structure, interpretation of social cognition, and explanation of social context. The study found the use of macrostructure, superstructure, and microstructure as textual elements that support the discourse theme. Meanwhile, the theme selection is influenced by the host’s ideology that serves the theme's pro-coalition stance. The social context underlying the discourse production is the representation of Joko Widodo, who utilized his power to forge a coalition with the opponents through political bargaining. The implication of this study is to increase public awareness on the importance of critically viewing political news so that the negative excesses of political propaganda do not easily provoke them through mass media.

Intonation structuring of coherent Tuvan folklore narration

The article presents preliminary data on the intonation of the Tuvan folklore narration with specific focus to the relation between intonation and text structure. The analysis was carried out on the basis of four Tuvan folklore texts. Three hypothetical intonational correlates of the text structure were examined. First, we consider the difference in tempo of the main parts of the texts, following the idea of three-part structure of folklore texts proposed by V. Propp (beginning, complication and ending). The data obtained show no direct correspondence between the tempo of an utterance and its position in the text, as the tempo of the first and the last utterances and their ratio to the average tempo vary significantly from text to text. Secondly, it is shown that the texts contain a number of metatext markers, which are used quite often and are distinguished by the means of intonation (including changes in tone and intensity and separation by a pause). Thirdly, verbal forms with particle -tyr performing similar functions are examined. In contrast to metatext markers, they are not characterized by any intonational prominence, as the tone and intensity follow the general line of declination marking the end of an utterance. Thus, intonation plays an important role in forming the structure of Tuvan folklore texts, which, however, manifests itself only indirectly, in the way of emphasizing lexical means of structuring the text (metatext markers).

Do EFL Teachers Reduce the Reading Gap in Qatar? A Study of Strategy Instruction in Government Schools

Teachers’ instructional practices are crucial to students’ achievement in reading comprehension. Students’ low English reading competence in Qatar’s government schools raises concerns about not only reading comprehension strategy instruction, but also, subsequently, learning as a natural outcome of teachers’ praxis. This quantitative survey investigated English as a foreign language (EFL) reading comprehension strategy instruction of a broad cross-sectional sample of (1-12) EFL teachers (n=754) of government schools in Qatar. The study examines teachers’ most frequently used strategies and explicit strategy instruction. To relate the findings of an exploratory data analysis, descriptive statistics, including means, standard deviations, and frequencies, were calculated for each 5-point Likert scale questionnaire item using the Statistical Package for the Social Sciences (SPSS-Version 24) statistics software. Results revealed that participant teachers reported a generally moderate and high frequency of using comprehension strategies (lowest mean 3.56 and highest mean 4.52). The seven most used strategies (mean score ≥4) were: Identify main ideas, set purpose for reading, predictions, preview text, monitor comprehension, prior knowledge activation, and handle unfamiliar words. Conversely, the five strategies reported a comparatively lower use frequency (mean score ˂4) were: Text structure, questioning, visualizing, summarizing, and think aloud. The major findings on explicit strategy instruction indicated that teachers ignored the gradual release of responsibility to students. The implications of these findings suggest that EFL teachers demonstrate moderate use of reading comprehension strategy instruction. Further recommendations for the Ministry of Education, school principals, and teachers are offered.