A PHP Error was encountered

Severity: 8192

Message: Non-static method URL_tube::usage() should not be called statically, assuming $this from incompatible context

Filename: url_tube/pi.url_tube.php

Line Number: 13

KDD 2020 | Order Fulfillment Cycle Time Estimation for On-Demand Food Delivery

Accepted Papers

Order Fulfillment Cycle Time Estimation for On-Demand Food Delivery

Lin Zhu: Alibaba Group; Wei Yu: Alibaba Group; Kairong Zhou: Alibaba Group; Xing Wang: Alibaba Group; Pengyu Wang: Alibaba Group; Wenxing Feng: Alibaba Group; Ning Chen: Alibaba Group; Pei Lee: Alibaba Group


By providing customers with conveniences such as easy access to an extensive variety of restaurants, effortless food ordering and fast delivery, on-demand food delivery (OFD) platforms have achieved explosive growth in recent years. A crucial machine learning task performed at OFD platforms is prediction of the Order Fulfillment Cycle Time (OFCT), which refers to the amount of time elapsed between a customer places an order and he/she receives the meal. The accuracy of predicted OFCT is important for customer satisfaction, as it needs to be communicated to a customer before he/she places the order, and is considered as a service promise that should be fulfilled as well as possible. As a result, the estimated OFCT also heavily influences planning decisions such as dispatching and routing.

In this paper, we present the OFCT prediction model that is currently deployed at Ele.me, which is one of the world’s largest OFD platforms and delivers over 10 million meals in more than 200 Chinese cities every day. By dissecting the order fulfillment cycle of a meal order, we identify key factors behind OFCT, and capture them with numerous features constructed using a wide range of data sources. These features are fed into a deep neural network (DNN), which further incorporates representations of couriers, restaurants and delivery destinations to enhance prediction efficacy. Finally, a novel post-processing layer is introduced to improve convergence speed by better accounting for the distributional mismatch between the true OFCT values and those predicted by the model at initialization. Extensive offline and online experiments demonstrate the effectiveness of our approach.

How can we assist you?

We'll be updating the website as information becomes available. If you have a question that requires immediate attention, please feel free to contact us. Thank you!

Please enter the word you see in the image below: