A Study on the Structure and Content of Abstracts: Focused on Journal of the Korean Society for Information Management

Chang, Woo-Kwon

doi:10.3743/kosim.2016.33.3.107

Cited by 1 publication

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The advantage of Scientometrics is that it can objectively and quantitatively measure the importance of a research topic through the frequency of keywords, and it is easy to grasp the size and history of the academic field to which the thesis belongs [9]. In the global information age, it is vital to quickly and accurately secure the information we need, so it is meaningful in terms of efficiency to analyze the abstract that notifies and criticizes the content of the thesis as an information retrieval tool [10]. In particular, if topic modeling is applied to the abstract, it is possible to comprehend all the words and messages appearing in the corresponding message by organizing the contents by topic [11].…”

Section: Introductionmentioning

confidence: 99%

Digital Health Discussion Through Articles Published Until the Year 2021: A Digital Topic Modeling Approach (Preprint)

Sung¹,

Kim²

2022

Preprint

View full text Add to dashboard Cite

BACKGROUND Since the 2010s, the digital health industry has grown significantly, gaining popularity with the public. The term “Digital Health” is being explored in various academic fields, such as public health, medicine, and computer science. OBJECTIVE This study analyzes the research trends of digital health related articles published in the Web of Science until 2021 to understand the research concentration, boundary, scope, and characteristics. METHODS By crawling and preprocessing 27,638 digital health-related papers provided by Web of Science and investigating 15,950 of them, the number of articles published by year and by field are compared and analyzed. Since these 15,950 papers belong to the top 10 academic fields, they were regrouped into three major fields: public health, medicine, and electrical engineering and computer science (EECS). Latent Dirichlet Allocation (LDA) is applied as a topic modeling method for each field and time period. The number of topics is determined based on the coherence score. RESULTS The number of optimal topics in the first and second halves for public health were 13 and 19, for medicine, 14 and 25, and for EECS, 7 and 21, respectively. Text analysis showed that articles from public health, medicine, and EECS share similar topics but vary in composition. The homogeneity test showed that the contrast between each group is significant (p<2.2e-16). All the topics revealed in articles could be categorized into six dominant themes; journal article methodology, information technology, medical issues, subject, social phenomenon, and healthcare. As a result of the LDA analysis, the topics of each domain differed, and the composition of each theme was different between academic fields and time periods. Studies on public health focused on social phenomena, prevention, and daily care, while studies in medicine investigated treatment and cure issues in the second half. Studies in EECS highlighted the importance of technical issues, while showing a comparatively distant relation to public health or medicine. All fields emphasized information technology (IT) in the first half, and each domain published specialized articles in the second half. In particular, there were numerous articles belonging to both public health and medicine, while only a few were common with EECS. CONCLUSIONS The articles belonging to each domain became more specialized and distinguished from other domains and all three fields highlighted social phenomena and healthcare over time. With Covid-19 becoming a dominant issue recently, digital health has come to be strongly related to depression and mental disorders, education, and physical activity with articles on these topics appearing in the second half in all fields. The scope of digital health research is expanding and its composition fluctuating. In the future, it will be necessary to explore papers on expanded topics that reflect people's needs for digital health.

show abstract

Section: Introductionmentioning

confidence: 99%