An e-commerce transaction record contains basic information such as the product, the user and the purchasing time. As the online shopping becomes popular, a large amount of transaction records have been continuously generated by users.

In this paper, we make use of the large volume of transaction records and products. We propose to detect and use the bursty patterns to analyze sales time series. So far, most existing studies use burst detection in text mining area, like study the news burst to detect new events. However, few have applied this method on e-commerce area.

We consider a problem of analyzing transaction records in both temporal and semantic domains, with the specific goal of identifying the best-seller and less-sold, aperiodic and periodic products. The transaction records of each product are transformed to a time series, where each element is the sales count at a certain point of time. With bursts detected, we categorize products into several types with the bursty feature. In addition, we also propose a method to use the the features to predict the future sales.

Keywords: e-commerce, time series, burst detection, sales prediction

[1] D. Agrawal, R. P. Agrawal, J. B. Singh and S. P. Tripathi, “E-commerce: True Indian Picture”, Journal of Advances in IT, vol. 3, no. 4, (2012), pp. 250-257.

[2] M. Vlachos, C. Meek, Z. Vagena, and D. Gunopulos. Identifying similarities, periodicities and bursts for online search queries. In Proc. 2004 ACM SIGMOD international conference on Management of data, pages 131–142, 2004.

[3] F Wu and B A Huberman. Novelty and collective attention. PNAS, 104:17599, 2007.

[4] E. Adar, D.S. Weld, Bershad, B.N., and S.D. Gribble. Why we search: visualizing and predicting user behavior. In Proc. WWW2007, pages 161–170, 2007.

[5] J.-P. Onnela and F. Reed-Tsochas. Spontaneous emergence of social influence in online systems. Proc. Natl Acad. Sci., 107:18375–18380, 2009.

[6] S. Goel, J. M Hofman, S. Lahaie, D. M Pennock, and D.J. Watts. Predicting consumer behavior with web search. PNAS, 107(41):17486–17490, 2010.

[7] J. Ratkiewicz, F. Menczer, S. Fortunato, A. Flammini, and A. Vespignani. Traffic in social media ii: Modeling bursty popularity. In SocialCom 2010: SIN, 2010.

[8] A. Kulkarni, Product Manager: http://yourstory.in/2013/01/indian-e-commerce-what-does-the-futurelook-like/.

[9] Avery, S. Online tool removes costs from process. Purchasing, vol. 123, no. 6, (1997), pp. 79-81.

[10] Internet Research: Electronic Networking Applications and Policy, vol. 8, no. 3, (1998), pp. 219-228.

[11] China Electronic Commerce Research Center, The 2012 annual China network retail market data monitoring report，July (2012)

[12] Ji SHu-xian, Zhao Dong-mei, Validity of the trust model in online reputation feedback based on time value. Journal of Applied statistics and management, vol.30, no. 6, (2011), pp. 1061-1066.

[13] August-WiLSelm Scheer, Absatzprognosen engl. Sales Forecasting, Springer Verlag, Berlin, 1983.

[14] Manfred Huttner, ¨ Markt- und Absatzprognosen engl. Market and Sales Forecasting, KoHMhammer, Stuttgart, 1982.

[15] Christian Abele, Michael Schaidnagel, Fritz Laux, Ilia Petrov, Sales Prediction with Parametrized Time Series Analysis, Proc. DBKDA 2013.

[16] Kleinbert, J.: Bursty and Hierarchical Structure in Streams. In: 8th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 91—101. ACM Press, Edmonton (2002)

[17] R. Kumar, J. Novak, P. Raghavan, and A. Tomkins. On the bursty evolution of blogspace. In WWW, pages 159–178, 2005.

[18] Q. Mei and C. Zhai. Discovering evolutionary theme patterns from text: an exploration of temporal text mining. In SIGKDD, pages 198–207, 2005.

[19] Q. He, K. Chang, E.-P. Lim, and J. Zhang. Bursty feature reprensentation for clustering text streams. In SDM, accepted, 2007.

[20] Fung, G.P.C., Yu, J.X., Yu, P.S., Lu, H.: Parameter Free Bursty Events Detection in Text Streams. In: 31st International Conference on Very Large Data Bases, pp. 181—192. ACM Press, Trondheim(2005).

[21] He, Q., Chang, K., Lim, E.P.: Analyzing Feature Trajectories for Event Detection. In: 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 207-214. ACM Press, Amsterdam(2007).

[22] Ratkiewicz et al., 2010, Leskovec et al., 2009, Crane and Sornette, 2008, Lehmann et al., 2012, Yang and Leskovec, 2011, Aral et al., 2009.

[23] Lehmann J., Goncalves B., Romasco J.J., Cattuto C. (2012), ‘Dynamical Classes of Collective Attention in Twitter’. In: Proccedings of the 21st International Conference on World Wide Web, WWW ’12, New York, NY, USA, ACM: 251-260.

[24] He, Q., Chang, K., Lim, E.P.: Analyzing Feature Trajectories for Event Detection. In: 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 207-214. ACM Press, Amsterdam(2007).