BPO数据生产平台的设计与优化

Design and optimization for BPO-data service common platform

根据对在线数据业务外包服务(data business process outsourcing,Data BPO)处理流程中存在的一系列瓶颈的描述和分析,提出了现有的BPO数据处理平台进行优化设计的方案,构建了BPO数据生产平台的优化模型。该模型有针对性地使用了包括自动、半自动模式识别技术,服务器集群动态负载均衡以及非对称网络传输等技术,有效地解决了数据生产平台的吞吐量、服务器负载以及网路传输瓶颈,提高了整个数据处理平台的效率及准确率。

Based on the descriptions and analyses for the bottle-necks in the current processing procedure of data business process outsourcing （BPO-Data）, a solution is presented and a model for processing BPO-Data is built optimally. In this model, some full auto or semi-auto pattern recognition technologies in image segment, dynamic load balance in server cluster and asymmetric network transmission will be combined into a solution to these bottle-necks which limit the throughput of data center, the pay- load capability of the server cluster and the capacity of network transmission. After the system optimization, the performance, including the throughput, response time and correction of the BPO-Data processing platform is promoted to a new high level.