نویسندگان
1 دانشگاه اصفهان؛ اصفهان، ایران.
2 دانشگاه کوروینوس بوداپست؛ بوداپست، مجارستان
3 دانشگاه یزد؛ یزد، ایران.
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
This research aims to verify the patterns of users' tagging to the articles of knowledge and information science field in academia.edu. The research method is quantitative and based on text mining and applicable typically. The population includes 6086 bibliographic articles and their abstracts extracted from 159 English journals of knowledge and information science field in Scopus database that are core journals in LISTA as well. In order to gather these data, 194337 articles were searched in academia.edu then every article that tagged was chosen. Examining the relationship between the growth of different types of tags (one-word, two-word, three-word, and four-word and more) and increasing of documents showed a linear correlation between them. Among the different groups of tags, the highest growth rate was related to two-word tags (.609%) and the lowest growth rate was related to four-word tags and more (.143%). The total growth rate of the tags (new and duplicate) was also 5.52 (i.e. 5.52 tags per document). It was also found that two-word tags had the most matching (54.92%) and four-word tags and the least matching (1.76%) with different sections of articles (title, abstract, and authors' keywords). The total tags were matched 7.5% with the title, 76.61% with the abstract, and 15.89% with the authors' keywords. Regarding the reuse of tags, it was revealed that in general, 38.8% of the tags have been reused. On the other hand, two-word tags with 57.59% had the most reuse and four-word tags and more with 7.54% had the least. Another point is that 16% of the tags were reused in the first year and more than 50% of the tags were reused in the first 3 years. Finally, it can be said that the existence of a significant user consensus on certain terms indicates that the new patterns of user tags are at least partially compatible with professional indexing concepts about document content, and by focusing on the most widely used tags and their sustainable distribution, weight formulation and even classification schemes may be achieved. Also, users' activities on social networks can be used to increase the quality of suggestions in collective tagging systems. Another point is that there is a connection between professional indexing and user tagging, and the two are not alien to each other. This connectivity can be the basis for a complementary subject access system that enriches professional indexing.
کلیدواژهها [English]
)