The environmental justice research on urban–rural exposure to flooding is underdeveloped and few empirical studies have been conducted in China. This study addresses this gap by exploring the probabilities of exposure to floods (10-, 20-, and 50-year) and examining the relationship between vulnerable groups and flooding in Nanjing, an important central city on the Yangtze River. Statistical analysis is based on multivariable generalised estimating equation (GEE) models that describe sociodemographic disparities at the census-tract level. The results revealed that (1) highly educated people in the urban centre are more likely to live in areas with high flood risk because of the abundance of education resources, and employment opportunities are concentrated in the urban centre. (2) Natives in suburban areas are more likely to live in flood-prone areas due to their favourable ecological environments near rivers and lakes. (3) Women in rural areas are more likely to live in high-flood-risk zones because most of the men are migrant workers. These findings highlight the urgent need to develop mitigation strategies to reduce flood exposure, especially in districts with high proportions of socially disadvantaged people. The linkages between rural and urban areas need to be strengthened in order to reduce flood exposure.

The identification of vulnerable people and places to flood is crucial for effective disaster risk management. Here, we combine flood hazard and social vulnerability index to capture the potential risk of flood. In this paper, Nanjing was taken as the case study to explore the spatial pattern of social vulnerability towards flood at the community scale by developing an index system. Based on the flood risk results of ArcSWAT, the risk of flood disaster in Nanjing was evaluated. The results show the following. (1) Social vulnerability exhibits a central–peripheral pattern in general, which means that the social vulnerability degree is high in the central city and decreases gradually to the suburbs. (2) The susceptibility to flood disaster has a similar circle-layer pattern that is the highest in the urban centre, lower in the exurban areas, and the lowest in the suburb areas. (3) By using the GIS-based zoning approach, communities are classified into four types by comprehensively considering their flood susceptibility and social vulnerability. The spatial pattern is explained, and policy recommendation for reducing flood risk is provided for each type of community. The research has important reference significance for identifying the spatial pattern of social vulnerability to flood and then formulating targeted adaptation countermeasures.

No abstract

In sponsored search, retrieving synonymous keywords for exact match type is important for accurately targeted advertising. Datadriven deep learning-based method has been proposed to tackle this problem. An apparent disadvantage of this method is its poor generalization performance on entity-level long-tail instances, even though they might share similar concept-level patterns with frequent instances. With the help of a large knowledge base, we find that most commercial synonymous query-keyword pairs can be abstracted into meaningful conceptual patterns through concept tagging. Based on this fact, we propose a novel knowledge-driven conceptual retrieval framework to mitigate this problem, which consists of three parts: data conceptualization, matching via conceptual patterns and concept-augmented discrimination. Both offline and online experiments show that our method is very effective. This framework has been successfully applied to Baidu's sponsored search system, which yields a significant improvement in revenue.