Papers › Multi-Turn Response Selection for Chatbots with Deep Attention Matching Network

Multi-Turn Response Selection for Chatbots with Deep Attention Matching Network

1 Jul 2018ACL 2018 7archive 2025-07-28

Xiangyang Zhou, Lu Li, daxiang dong, Yi Liu, Ying Chen, Wayne Xin Zhao, dianhai yu, Hua Wu

Human generates responses relying on semantic and functional dependencies, including coreference relation, among dialogue elements and their context. In this paper, we investigate matching a response with its multi-turn context using dependency information based entirely on attention. Our solution is inspired by the recently proposed Transformer in machine translation (Vaswani et al., 2017) and we extend the attention mechanism in two ways. First, we construct representations of text segments at different granularities solely with stacked self-attention. Second, we try to extract the truly matched segment pairs with attention across the context and response. We jointly introduce those two kinds of attention in one uniform neural network. Experiments on two large-scale multi-turn response selection tasks show that our proposed model significantly outperforms the state-of-the-art models.

PaperPDFCode

Code

baidu/Dialogue officialmentioned in papertf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

ChatbotConversational Response SelectionTranslation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Conversational Response Selection Douban DAM MAP 0.550 #15 of 16 Archive leaderboard report
Conversational Response Selection Douban DAM MRR 0.601 #15 of 16 Archive leaderboard report
Conversational Response Selection Douban DAM P@1 0.427 #15 of 16 Archive leaderboard report
Conversational Response Selection Douban DAM R10@1 0.254 #15 of 16 Archive leaderboard report
Conversational Response Selection Douban DAM R10@2 0.410 #15 of 16 Archive leaderboard report
Conversational Response Selection Douban DAM R10@5 0.757 #15 of 16 Archive leaderboard report
Conversational Response Selection RRS DAM MAP 0.511 #6 of 7 Archive leaderboard report
Conversational Response Selection RRS DAM MRR 0.534 #6 of 7 Archive leaderboard report
Conversational Response Selection RRS DAM P@1 0.347 #6 of 7 Archive leaderboard report
Conversational Response Selection RRS DAM R10@1 0.308 #6 of 7 Archive leaderboard report
Conversational Response Selection RRS DAM R10@2 0.457 #6 of 7 Archive leaderboard report
Conversational Response Selection RRS DAM R10@5 0.751 #6 of 7 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) DAM R10@1 0.767 #19 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) DAM R10@2 0.874 #19 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) DAM R10@5 0.969 #19 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) DAM R2@1 0.938 #19 of 25 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerReLUResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections