Mining New Business Opportunities: Identifying Trend related Products by Leveraging Commercial Intents from Microblogs

Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, pages 1337–1347, Seattle, Washington, USA, 18-21 October 2013. c⃝2013 Association for Computational Linguistics Mining New Business Opportunities: Identifying Trend related Products by Leveraging Commercial Intents from Microblogs Jinpeng Wang1, Wayne Xin Zhao1, Haitian Wei2, Hongfei Yan1 and Xiaoming Li1 1School of Electronic Engineering and Computer Science, Peking University, China 2Daton Securities Co., Ltd., No.93 Jianguo Rd, Chaoyang District, Beijing, China JooPoo@pku.edu.cn, {batmanfly,haataa.wei,yhf1029}@gmail.com, lxm@pku.edu.cn Abstract Hot trends are likely to bring new business opportunities. For example, “Air Pollution” might lead to a significant increase of the sales of related products, e.g., mouth mask. For e- commerce companies, it is very important to make rapid and correct response to these hot trends in order to improve product sales. In this paper, we take the initiative to study the task of how to identify trend related products. The major novelty of our work is that we au- tomatically learn commercial intents revealed from microblogs. We carefully construct a da- ta collection for this task and present quite a few insightful findings. In order to solve this problem, we further propose a graph based method, which jointly models relevance and associativity. We perform extensive experi- ments and the results showed that our methods are very effective. 1 Introduction A trend is a hot topic (e.g., the release of a popular movie) which is being widely discussed by the pub- lic. Hot trends usually attract much attention from the public, and they are likely to bring new business opportunities. Consider the following e-shopping s- cenario. A user in Beijing would like to buy some- thing to reduce the health impacts of Beijing air pol- lution1. Different from traditional e-shopping sto- ries, in this case the user may not have a clear idea of what she should buy, and cannot even formulate the 1See http://www.nytimes.com/2013/04/04/world/asia/two-major- air-pollutants-increase-in-china.html to find more details about the trending topic “Beijing Air Pollution”. purchase needs into a clear query. Faced with trend- driven business opportunities, e-commerce compa- nies typically ask workers to manually identify relat- ed products and make heuristic rules to match user queries (e.g., incorporating trending keywords into related product titles). To improve trend-driven e-commerce, in this pa- per, we propose to study the novel task of automat- ically identifying trend related products. Why is it compelling to understand and study trend-driven product purchase? Because hot trends are closely related to business opportunities directly or indirect- ly. As a case of direct causal relationship, the world- wide popularity of the movie series “Harry Potter” created the great success of the original novels of “Harry Potter”. As a case of indirect causal rela- tionship, the stock rise or salary increase might exert positive effects on product sale. Based on our empir- ical analysis (See Section 3), a considerable propor- tion, i.e. 50%, of hot trends discussed on the largest Chinese microblog (i.e. Sina Weibo) indeed have corresponding product entries in the largest Chinese C2C e-commerce website (i.e. Taobao), which …

Content truncated to 500 words