STUDY OF CYCLIC TRADING PATTERN ON LARGE NETWORK OF UNIFIED INVOICE

Unified invoice was uniquely introduced in Taiwan in early 1950s, which is a form of state lottery managed by the Financial Information Agency of Taiwan and was literally used as an incentive to boost tax revenues. In its first launch, the tax revenue increased to 75% from that of the previous year. Its effectiveness leads to a mandate to use the unified invoice as a standard way to record every domestic business transaction. The introduction of electronic and cyber technologies facilitates the management and operation of the practices of the unified invoice. There are approximately 140 million records of transaction accumulated per year, which form a trading network. An understanding of the structure of the trading network might be able to provide us a reference to investigate trading behaviors in Taiwan both in macro and micro scale. In this study, a sequence of approximately 400+ million transaction records, from 2010 to 2012, excluding the transactions records from the listed companies in the stock markets, which behavior is well regulated and monitored. We explore with the typical network analysis techniques and then extend it with the persistent homology of the trading network, i.e., topology of cyclic trading pattern. The evolution of the trading network based on the order of the sequence proves useful to understand the lifespan of trading behavior of the individual company. The initial result can be used to develop tools for anomalies detection and identify possible specific local activities.

關鍵字

Unified Invoice ； Trading Network ； Cyclic Pattern ； Persistent Homology

參考文獻

Akoglu, L., Tong, H., & Koutra, D. (2015). Graph based anomaly detection and description: A survey. Data Mining and Knowledge Discovery, 29, 626-688. doi:10.1007/s10618-014-0365-y

Bank, R. E., & Douglas, C. C. (1993). Sparse matrix multiplication package (SMMP). Advances in Computational Mathematics, 1, 127-137. doi:10.1007/BF02070824

Bhawiyuga, A., & Kirana, A. P. (2016, July). Implementation of page rank algorithm in Hadoop MapReduce framework. Paper presented at the 2016 International Seminar on Intelligent Technology and Its Applications, Lombok, Indonesia. doi:10.1109/ISITIA.2016.7828663

Chang, W. L., Grady, N., & NBD-PWG NIST Big Data Public Working Group. (2015). NIST big data interoperability framework: Volume 1, definitions. doi:10.6028/NIST.SP.1500-1

Edelsbrunner, H., Letscher, D., & Zomorodian, A. (2000, November). Topological persistence and simplication. Paper presented at the 41st Annual Symposium on Foundations of Computer Science, Redondo Beach, CA. doi:10.1109/SFCS.2000.892133

國際替代計量

STUDY OF CYCLIC TRADING PATTERN ON LARGE NETWORK OF UNIFIED INVOICE

全文下載

主題瀏覽