Hierarchical Attention Network for Context-Aware Query Suggestion

Li, Xiangsheng; Liu, Yiqun; Li, Xin; Luo, Cheng; Nie, Jian‐Yun; Zhang, Min; Ma, Shaoping

doi:10.1007/978-3-030-03520-4_17

Cited by 3 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The click-through bipartite graph specifies co-clicked documents with respect to a given query, which reduces the lexical gap between query and document embeddings. Compared to previous neural ranking models, the learned representations can better model users' search intent, which is particularly important for long-tail queries [15,18].…”

Section: Introductionmentioning

confidence: 99%

Learning Better Representations for Neural Information Retrieval with Graph Information

Rijke

Liu

et al. 2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

Self Cite

View full text Add to dashboard Cite

Neural ranking models have recently gained much attention in information retrieval (IR) community and obtain good ranking performance. However, most of these retrieval models focus on capturing the textual matching signals between query and document but do not consider user behavior information that may be helpful for the retrieval task. Specifically, users' click and query reformulation behavior can be represented by a click-through bipartite graph and a session-flow graph, respectively. Such graph representations contain rich user behavior information and may help us better understand users' search intent beyond the textual information. In this study, we aim to incorporate this rich information encoded in these two graphs into existing neural ranking models.We present two graph-based neural ranking models (EmbRanker and AggRanker) to enrich learned text representations with graph information that captures rich users' interaction behavior information. Experimental results on a large-scale publicly available benchmark dataset show that the two models outperform most existing neural ranking models that only consider textual information, which illustrates the effectiveness of integrating graph information with textual information. Further analyses show how graph information complements text matching signals and examine whether these two models can be adopted in practical applications.

show abstract

Section: Introductionmentioning

confidence: 99%