Automatic Title Generation for Learning Resources and Pathways with Pre-trained Transformer Models

2021 ◽  
Vol 15 (04) ◽  
pp. 487-510
Author(s):  
Prakhar Mishra ◽  
Chaitali Diwan ◽  
Srinath Srinivasa ◽  
G. Srinivasaraghavan

To create curiosity and interest for a topic in online learning is a challenging task. A good preview that outlines the contents of a learning pathway could help learners know the topic and get interested in it. Towards this end, we propose a hierarchical title generation approach to generate semantically relevant titles for the learning resources in a learning pathway and a title for the pathway itself. Our approach to Automatic Title Generation for a given text is based on pre-trained Transformer Language Model GPT-2. A pool of candidate titles are generated and an appropriate title is selected among them which is then refined or de-noised to get the final title. The model is trained on research paper abstracts from arXiv and evaluated on three different test sets. We show that it generates semantically and syntactically relevant titles as reflected in ROUGE, BLEU scores and human evaluations. We propose an optional abstractive Summarizer module based on pre-trained Transformer model T5 to shorten medium length documents. This module is also trained and evaluated on research papers from arXiv dataset. Finally, we show that the proposed model of hierarchical title generation for learning pathways has promising results.

Author(s):  
Htay Htay Win ◽  
Aye Thida Myint ◽  
Mi Cho Cho

For years, achievements and discoveries made by researcher are made aware through research papers published in appropriate journals or conferences. Many a time, established s researcher and mainly new user are caught up in the predicament of choosing an appropriate conference to get their work all the time. Every scienti?c conference and journal is inclined towards a particular ?eld of research and there is a extensive group of them for any particular ?eld. Choosing an appropriate venue is needed as it helps in reaching out to the right listener and also to further one’s chance of getting their paper published. In this work, we address the problem of recommending appropriate conferences to the authors to increase their chances of receipt. We present three di?erent approaches for the same involving the use of social network of the authors and the content of the paper in the settings of dimensionality reduction and topic modelling. In all these approaches, we apply Correspondence Analysis (CA) to obtain appropriate relationships between the entities in question, such as conferences and papers. Our models show hopeful results when compared with existing methods such as content-based ?ltering, collaborative ?ltering and hybrid ?ltering.


2020 ◽  
Vol 24 (6) ◽  
pp. 1311-1328
Author(s):  
Jozsef Suto

Nowadays there are hundreds of thousands known plant species on the Earth and many are still unknown yet. The process of plant classification can be performed using different ways but the most popular approach is based on plant leaf characteristics. Most types of plants have unique leaf characteristics such as shape, color, and texture. Since machine learning and vision considerably developed in the past decade, automatic plant species (or leaf) recognition has become possible. Recently, the automated leaf classification is a standalone research area inside machine learning and several shallow and deep methods were proposed to recognize leaf types. From 2007 to present days several research papers have been published in this topic. In older studies the classifier was a shallow method while in current works many researchers applied deep networks for classification. During the overview of plant leaf classification literature, we found an interesting deficiency (lack of hyper-parameter search) and a key difference between studies (different test sets). This work gives an overall review about the efficiency of shallow and deep methods under different test conditions. It can be a basis to further research.


2021 ◽  
Vol 15 (6) ◽  
pp. 1-22
Author(s):  
Yashen Wang ◽  
Huanhuan Zhang ◽  
Zhirun Liu ◽  
Qiang Zhou

For guiding natural language generation, many semantic-driven methods have been proposed. While clearly improving the performance of the end-to-end training task, these existing semantic-driven methods still have clear limitations: for example, (i) they only utilize shallow semantic signals (e.g., from topic models) with only a single stochastic hidden layer in their data generation process, which suffer easily from noise (especially adapted for short-text etc.) and lack of interpretation; (ii) they ignore the sentence order and document context, as they treat each document as a bag of sentences, and fail to capture the long-distance dependencies and global semantic meaning of a document. To overcome these problems, we propose a novel semantic-driven language modeling framework, which is a method to learn a Hierarchical Language Model and a Recurrent Conceptualization-enhanced Gamma Belief Network, simultaneously. For scalable inference, we develop the auto-encoding Variational Recurrent Inference, allowing efficient end-to-end training and simultaneously capturing global semantics from a text corpus. Especially, this article introduces concept information derived from high-quality lexical knowledge graph Probase, which leverages strong interpretability and anti-nose capability for the proposed model. Moreover, the proposed model captures not only intra-sentence word dependencies, but also temporal transitions between sentences and inter-sentence concept dependence. Experiments conducted on several NLP tasks validate the superiority of the proposed approach, which could effectively infer meaningful hierarchical concept structure of document and hierarchical multi-scale structures of sequences, even compared with latest state-of-the-art Transformer-based models.


1986 ◽  
Vol 49 ◽  
pp. 1-4
Author(s):  
Thomas E. Cronin

In recent years I've supervised countless undergraduate research papers and numerous senior theses. Not surprisingly, I repeat myself about basic research and writing hints, suggestions, and outright warnings. This handout, “Write Tigers Write!” prepared for the “tigers” at Princeton and The Colorado College (the mascot is the same), attempts to help the novice researcher and is, as well, an act of self-protection.What follows are suggestions and cautions for students writing a research paper. My suggestions are merely that. They are personal, general and speak more about writing than about research.Make no mistake about it. Research and writing are demanding work even for the professional. You won't hear professional scholars or writers boast about the easiness of their craft. No matter how much they love it, and they often love it more than anything else, they find it demanding, exacting, lonely and often painful—if they really work at it.


2014 ◽  
Vol 9 (2) ◽  
pp. 37
Author(s):  
Andrea Baer

A Review of: Schwegler, R. A., and Shamoon, L. K. (1982). The aims and process of the research paper. College English, 44(8), 817-824. Objectives – This classic article discusses research-based writing assignments. Schwegler and Shamoon sought to identify differences between college students’ and college instructors’ conceptions of research and research paper assignments, particularly in terms of their purpose and process. The authors also sought to identify common features of academic research writing that could inform writing instruction about research writing. Design – Qualitative interviews with college instructors and students about their views of the research process and about forms of research writing. Instructors were also interviewed about evaluation standards for academic research papers. Setting – Unspecified, though the description suggests a college or university in the United States. Subjects – College instructors and college students. (Number of subjects unspecified.) Methods – The authors, a university writing program director and a writing program instructor, conducted one-on-one interviews with college instructors and students about their views of research and the research paper. Questions focused on conceptions of the research process, the purposes of research, and the forms that research writing takes. Instructors were also asked about standards for effective evaluation of research papers. The limited description of the research methods and interview questions employed in this study hinder the ability to critically assess its validity and reliability. Potential limitations of the study, such as selection bias or unclear wording of interview questions, cannot be adequately assessed based on the provided information. The authors also do not identify limitations of their study. As is discussed in more detail in this review’s commentary, the study does not conform to the conventions of most research studies from the behavioral, health, physical, and social sciences. The authors’ methods, however, may be better understood in light of particular disciplinary approaches and debates in Composition Studies. Main Results – Interviewees’ responses illustrated notable differences between college instructors’ and college students’ conceptions of the process, purpose, forms, and audiences of research paper assignments. While instructors understood the research paper to be argumentative, analytical, and interpretive, students generally described it as informative and factual. Students, when asked why research papers are assigned, identified purposes such as learning more about a topic, demonstrating one’s knowledge, or learning to use the library. Instructors indicated that the purpose of the research paper includes testing a theory, building on previous research, and exploring a problem that has been presented by other research or events (p. 819). At the same time, most instructors described research as an ongoing pursuit of “an elusive truth” (p. 819), rather than as primarily factual in nature. According to Schwegler and Shamoon, instructors also indicated during interviews that research and writing involve a clear though complex pattern that is evident in the structure and conventions of research papers. For example, the research process usually begins with activities like reading, note-taking, identifying problems with and gaps in current research, and conversing with colleagues. These instructors also reported that writing conventions which are implicitly understood in their fields are used by other scholars to evaluate their peers’ work. Reflecting on these interview responses, Schwegler and Shamoon suggest that pedagogical approaches to writing instruction can be informed both by acknowledging disparities in students’ and instructors’ conceptions of research and by identifying shared characteristics of academic writing. The authors therefore make several general observations about the nature of professional research papers and describe the structure and conventions of academic research papers. They conclude that the structure of scholarly research papers across the disciplines reflects the research process. Such a paper opens with identification of a research problem and a review of current knowledge and is followed by a variation of four possible patterns: 1) Review of research, 2) Application or implementation of a theory, 3) Refute, refine, or replicate prior research, and 4) Testing a hypothesis ( pp. 822-823). Schwegler and Shamoon indicate that the key features of scholars’ writings are also apparent in student research papers which instructors evaluate as highly-ranked and absent in lower-ranked papers. Furthermore, they provide an appendix that outlines the essential textual features of a research paper (Appendix A) (p. 822). It is unclear, however, if these descriptions of scholarly research writing are based on the instructor interviews or on other sources, such as previous analytical studies or an analysis of academic research papers from various disciplines. The researchers do not articulate the specific methods used to arrive at their generalizations. Conclusion – The authors conclude that students’ and instructors’ differing conceptions of the research process and the research paper have important implications for writing instruction. Many of the interviewed instructors described research as involving methods that are quite different from those needed for most research paper assignments. The discrepancies between class assignments and academics’ approaches to research suggests that differences in instructors’ and students’ views of research often are not addressed in the design of research paper assignments. Instructors who teach the research paper should ensure that the purpose, structure, and style of assignments reflect what content-area instructors will expect from students. Schwegler and Shamoon argue that because the basic conventions of the research paper generally apply across disciplines, instruction about those conventions can be integrated into composition courses and lower-level undergraduate courses. Such an approach can assist students in better understanding and approaching research writing as would a scholar in the given discipline.


1983 ◽  
Vol 19 (1) ◽  
pp. 17-24
Author(s):  
Melanie Vickers

Writing a research paper, if broken down into clearly defined steps, will enhance communication skills and feelings of success.


2021 ◽  
Vol 10 (01) ◽  
pp. 395
Author(s):  
Muhamad Uyun ◽  
Idi Warsah

The transformation of learning from offline to online modes due to the Covid-19 pandemic has affected students’ self-endurance in learning and the complexities of learning problems. Accordingly, this study was oriented towards investigating IAIN Curup students’ self-endurance and the problems they faced during online learning. Grounded in a qualitative method, this study incorporated 20 IAIN Curup students selected purposively as the participants. They were interviewed on the basis of the study’s foci. The data of interviews were analyzed using an interactive model. The findings demonstrated weak students’ self-endurance in online learning. Psychologically, they felt under pressure. Physically, their eyes’ condition got worse, and a few of them suffered from headache. They were psychologically bored as well less-concentrated, and they could not enjoy learning due to the limitation of social engagement in learning. Subsequently, as regards the problems faced during online learning, they had problems related to Internet’s signals, budgeting for fulfilling Internet’s needs, and difficulty in the search of learning resources, wherein the foregoing was categorized as a case of weak technological competence.


2018 ◽  
Vol 11 (10) ◽  
pp. 86
Author(s):  
Ranitha Sachinthana Weerarathna ◽  
Sharmini Perera

The main objective of this research paper to conceptualize a model to test the association between relationship conflicts and employees’ intention to leave with the support of past theories and research findings. This research paper contains the definitions for variables related to relationship conflicts and employees’ intention to leave from the organization and also it explains models which explain the interconnections among two variables. The conceptualization of this research includes a logically developed model that leads to identify the relationships between the independent variable and dependent variable. The proposed model suggested that there is an association between relationship conflicts and employees’ intention to leave.


Sign in / Sign up

Export Citation Format

Share Document