빅데이터 분석을 활용한 유통 분야 연구동향 분석

초록

Changes in the society and culture, as a result of technological development have a profound effect on academic development. It is crucial to systematically organize previous achievements in the past to adapt to such changes. In particular, considering that the 4th Industrial Revolutionary Technologies are expected to be applied in the field of distribution, it is expected that discussions in the field of distribution industry and academics will be growing. More systematic and detailed road-map is needed to reduce risks and uncertainty of long-term approach and research. In this study, by looking at the past and present research results in the distribution field, it is intended to examine to what extent and in what aspects the discussions on distribution has discussed. For the understanding of trends in the research, usually variables were set based on the classification criteria of previous studies, and then the existing documents were collected and read directly, and a content analysis method was used to check the frequency. This content analysis can provide deeper insights based on researcher’s knowledge. However, the subjectivity of the researcher can be involved and it is difficult to understand the relationship between key-words. To compensate this problem, it is necessary to use big data analysis methods. In particular, the big data analysis method is useful for quantitatively analyzing large amounts of data contained in literary materials such as academic papers, articles, and patents. There are several research manner through the big data analysis method. The one is to select a specific field and academic papers of specific field’s journals, The other is to select a specific keyword and analyze the literature related to the keyword. In this study, we will select a journal that represents the distribution field and analyze the literature accordingly to examine research trend. Previous studies looked at organizational buyer sector so called B2B, B2C, trends and insights on specific topics, and frequency analysis based on year, subject, numbers of publication, and number of researchers using content analysis method. However, those studies have not suggested new criteria for classification and only provided general description of previous researches. There are some studies using big data analysis method through text mining in the distribution field. However, deep insights are difficult to get, since the analysis were conducted only on the title of researches. Therefore, in this study, not only title, but also abstract, and key-words were analyzed. Also, diverse big data analysis methods such as word cloud, semantic network, and topic modeling were used. Word cloud analysis was implemented to analyze the frequency of certain key-words avoiding simple duplication of key-words, The result shows that the term ‘Relationship’ appeared in 392 academic papers, as the most frequent word, followed by other key-words such as ‘Distribution’, ‘Consumer’, ‘Enterprise’, and ‘Customer’. Topic modeling is based on the assumption that words in a document represent a document, and extracts the subject based on the probability of word distribution. As a result of topic modeling analysis, a total of 4 topics were drawn. Each topic can be classified into topics of channels, members, services, and organization and relationships. Through discussions among researchers, the topic names were given as topic path structure, distribution path members, distribution services, and distribution relationship management. Semantic network analysis is a method that examines the characteristics of the network through each entity and describes the network based on the relationship. The degree of relationship between key-words is judged as Degree Centrality, Closeness Centrality, Betweenness Centrality and Eigenvector Centrality. As a result of analysis, key-words differ in ranking, but key-words such as relationship, consumer, business, customer, and distribution appeared as an important keyword. In particular, 'relationship' and 'consumer' were very high. The results show that, first, academic papers in distribution research can be categorized into ‘relationship’, ‘consumer’, ‘enterprise’, ‘market’, and ‘customer’. Second, the topics in the distribution field were named as ‘Distribution Path Structure’, ‘Distribution Path Member’, ‘Distribution Service’, and ‘Distribution Relationship Management’, respectively. Third, research in the field of distribution forms a semantic network centered on keywords ‘consumption’ and ‘relationship’ and was classified into three clusters: ‘consumption’, ‘relationship’ and ‘franchise’. The implications of this study are as follows: First, this study cover the overall content of the field, so it is possible to examine how far researches on certain topic has progressed. Second, it has academic significance in that it is a new approach to the literature research method, and it can supplement existing content analysis methods. Third, the societal and industrial changes related to distribution is rapidly happening with the emergence of new technologies, but key-words related to it have not appeared in importance. Therefore, the scope of academic research topics in the distribution field is limited and various research topics are needed to be studied. The future research directions according to the limitations of this study are as follows. First, this study selected one journal and analyzed. In future research, the analysis can be conducted on plural journals and even certain key-words in all academic field can be analyzed. Second, this study has limitations in that it analyzes concentrated data such as titles, subject words, and abstracts, not al data. Therefore, if the entire text is included in the subject of analysis and discussed, deeper and more diverse insights will be possible Lastly, since this study analyzed data only on academic papers, there are n limitations in providing implications for practical application. Therefore, if analysis on practical data such as news articles and SNS, it will be able to produce more insightful and practical results.

키워드