基于社交媒体的商业区选址研究

Research on Commercial District Location Based on Social Media Data

针对利用传统的问卷调查数据对商圈选址存在样本小、效率低的问题,提出基于新浪微博签到数据,利用哈夫模型,以商业区的营业面积、顾客到商业区的路网距离与聚类后的顾客签到数据为输入,商业区的营业面积与顾客到商业区的路网距离的敏感参数为输出,得到用户的光临概率,继而用于商业区竞争力大小的计算,得到最佳的商业区选址。以北京五大商业区为例,得到校正后的哈夫模型的均方差小于15%,证明了所提方法的有效性。

Traditionally, the sample of the traditional questionnaire survey data to locating a new commercial district is small and the efficiency is low. Therefore, we put forward a method to locate a new commercial district based on social media check-in data. Using the Huff model, we took the area of commercial district, the network distance from demand point to facility and clustered customer check-in data as input and the sensitive parameters of area and network distance as output information, and obtained the probability plane of customer, which was used to estimate the competitiveness of the new commercial district. Finally, we achieved a best location of the new new commercial district. Taking the five big commercial districts in Beijing for example, we got the RMSE of parameters estimated values was less than 15%, which demonstrated the efficiency of the proposed method.