A journal article is consisted of the various data elements such as title, author, affiliation, abstract, keywords, tables, figures and references. It is required for authors to follow these items strictly as prescribed by each journal to write a pape...
A journal article is consisted of the various data elements such as title, author, affiliation, abstract, keywords, tables, figures and references. It is required for authors to follow these items strictly as prescribed by each journal to write a paper to submit. Thus, the detailed knowledge of the structure of a journal article is very important for the system developers. As how the structure of the journal article is reflected and indexed in database make critical effect on the capability of using the resources and the research, it is necessary to make the structural approach of the article and identify the data elements for the construction of effective database and the development of system.
In this study, 136 journals covered in Synapse were analyzed for each data element defined in the Journal Article Tag Suits(JATS), NISO Z39.96-201x. 8 key data elements such as author, affiliation, abstract, keywords, body, table, figure and reference were selected for this study and the 24 items were selected from 8 information factors as they show the distinctive characteristics for each journal and analyze differences of description styles of data elements for each journal. The findings are as follows:
A great variances in style for each element is identified in this study. In stating author names, most journals (128 journals, 94.1%) put first names first followed by last name, but 8 journals (5.9%) have names in "last name, first name" order. Author's degree is used in journals (33 journals, 24.3%). To indicate who the corresponding author is usually label is used in journals (111 journals, 81.6%). ORCID is adopted in journals (15 journals, 11.0%) as an author identification code. Affiliation label to match with author names are also various: numbering (121 journals, 89.0%); symbol (12 journals, 8.8%); alphabet (3 journals, 2.2%). And sequence of giving the label to affiliation is also various: from the first affiliation (116 journals, 85.3%); from the second affiliation (20 journals, 14.7%). In stating affiliations, the range of affilation is also various: city, country name (122 journals, 89.7%); city, country, post code (11 journals, 8.1%); only institution's name (3 journals, 2.2%). And 97 journals (71.3%) omit the repeated information in affiliation. In stating original article's abstract, there are two kinds of abstract: structured abstract (117 journals, 86.0%); unstructured abstract (19 journals, 14.0%). The structure of abstract is also various: 4-heading format (114 journals, 97.4%); 5-heading format (1 journal, 0.9%); 6-heading format (2 journals, 1.7%). Most journals (95 journals, 69.9%) limit to 250 words of abstract, and 22 journals (16.2%) limit to 300 words, 12 journals (8.8%) limit to 200 words, 5 journals (3.7%) limit to 350 words, and 2 journals (1.5%) limit to 400 words. In stating keywords, sign that separate keywords and sequence of list keywords are various. Most journals (87 journals, 64.0%) use comma, but semi colon (38 journals, 27.9%), middle dot(8 journals, 5.9%), slash(2 journals, 1.5%), and new line (1 journal, 0.7%) are also used. Most journal (111 journals, 81.6%) list keyword in the order of importance, whereas 25 journals list keywords in alphabetical order.
In main text, the styles of level of headings are various: all capital (53 journals, 39.0%), numbering (4 journals, 2.9%), coloring (48 journals, 35.3%) in first heading; all capital (2 journals, 1.5%), numbering (37 journals, 27.2%), coloring (19 journals, 14.0%), italic (4 journals, 2.9%) in second heading. In stating cited reference number, there are various descriptive style: [baseline] (63 journals, 46.3%); superscript (34 journals, 25.0%); superscript) (30 journals, 22.1%); (baseline) (9 journals, 6.6%). In stating cited figure in main text, Most journals (109 journals, 80.1%) abbreviate as "Fig." and 4 journals (2.9%) don't abbreviate it.
The sign used in table's footnote is also various: symbol (79 journals, 58.1%); alphabet (48 journals, 35.3%); number (4 journals, 2.9%). In figure, most journals (100 journals, 73.5%) abbreviate figure as "Fig." in label, whereas 36 journals (26.5%) don't abbreviate it. There are difference descriptive styles in references. The limit number of co-authors is various: 6(if>6) (86 journals, 63.2%); 3(if>6) (22 journals, 16.2%); all authors without limitation (20 journals, 14.7%); 3(if>3) (5 journals, 3.7%); 10(if>10) (2 journals, 1.5%); 3(if>4) (1 journal, 0.7%). Most journals(133 journals, 97.8%) write abbreviated journal title in reference and only 3 journals (2.2%) write full journals title in reference. In stating cited reference, only 2 journals (1.5%) write publication month and only 17 journals (12.5%) write issue number. And in stating last page of reference, most journals (88 journals, 64.7%) omit the repeated page number and 48 journals (35.3%) write the last page in full. Only 5 journals (3.7%) write DOI of the reference.
This study covers not only the references and abstracts but also the key data elements such as the full text index database to analyze the characteristics of each factor in order to find any distinctiveness. The results are expected to be an reliable source for the construction of effective search system and database.