Technology Map: A Text Mining and Network Analysis Approach

Abstract

The

Mapping and predicting research trends in international journal publications using graph and topic modeling

“…The graph is a set G={V, E}, where V is the set of nodes V={v1, v2, v3,..., vi} and E is the set of edges E={v1v2, v1v3,..., vivj}. In this research, V denotes the analyzed documents, for E is the relationship between two documents assessed by their association values [21]. Because relationships are visualized as graphs, network analysis can display crucial information from research graphs and simulate research growth activities.…”

“…𝒳 𝑗 is the eigenvector centrality of node j and λ constant value [43]. 7), where Iij is a binary value that has a value of 1 if research is a breakthrough from research j and 0 if research is incremental from research j [21].…”

“…This type of research/innovation is usually divided into two major parts, breakthrough and incremental. The research included in the breakthrough group is research without prior research; on the other hand, the research included in the incremental group is a development/continuation research from previous research in a research group [21]. The definition and scope adopted by this research are to distinguish preliminary research and further research so that the series of research developments are known.…”

“…First, Yoon and Park [8] use text mining and network analysis to create patent networks that show high-tech trends and identify new product development avenues. Second, Hakim et al [21] make a technology map in Indonesian-language journals using text mining and network analysis. Third, Abuhay et al [9] predict trends in scientific research topics in proceedings using non-negative matrix factorization (NMF) topic modeling and ARIMA prediction methods.…”

IPOC: an efficient approach for dynamic association rule generation using incremental data with updating supports

“…Deng and Lv [26], titled "PrePost+: An efficient N-lists-based algorithm for mining frequent itemsets via Children-Parent Equivalence pruning", wrote an algorithm named PrePost+ for efficient mining process. A tree was generated with left and right child's by make use of N-list as anetwork [27], and applied pruning at left part and right part of the tree referred as equivalence pruning. It proved more efficient than existing approaches.…”