Papers › Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots

Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots

7 Apr 2020arXiv:2004.03588archive 2025-07-28

Jia-Chen Gu, Tianda Li, Quan Liu, Zhen-Hua Ling, Zhiming Su, Si Wei, Xiaodan Zhu

In this paper, we study the problem of employing pre-trained language models for multi-turn response selection in retrieval-based chatbots. A new model, named Speaker-Aware BERT (SA-BERT), is proposed in order to make the model aware of the speaker change information, which is an important and intrinsic property of multi-turn dialogues. Furthermore, a speaker-aware disentanglement strategy is proposed to tackle the entangled dialogues. This strategy selects a small number of most important utterances as the filtered context according to the speakers' information in them. Finally, domain adaptation is performed to incorporate the in-domain knowledge into pre-trained language models. Experiments on five public datasets show that our proposed model outperforms the present models on all metrics by large margins and achieves new state-of-the-art performances for multi-turn response selection.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

JasonForJoy/SA-BERT officialmentioned in papermentioned on GitHubtf 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

Conversational Response SelectionDisentanglementDomain AdaptationRetrieval

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Conversational Response Selection Douban SA-BERT MAP 0.619 #8 of 16 Archive leaderboard report
Conversational Response Selection Douban SA-BERT MRR 0.659 #8 of 16 Archive leaderboard report
Conversational Response Selection Douban SA-BERT P@1 0.496 #8 of 16 Archive leaderboard report
Conversational Response Selection Douban SA-BERT R10@1 0.313 #8 of 16 Archive leaderboard report
Conversational Response Selection Douban SA-BERT R10@2 0.481 #8 of 16 Archive leaderboard report
Conversational Response Selection Douban SA-BERT R10@5 0.847 #8 of 16 Archive leaderboard report
Conversational Response Selection E-commerce SA-BERT R10@1 0.704 #8 of 15 Archive leaderboard report
Conversational Response Selection E-commerce SA-BERT R10@2 0.879 #8 of 15 Archive leaderboard report
Conversational Response Selection E-commerce SA-BERT R10@5 0.985 #8 of 15 Archive leaderboard report
Conversational Response Selection RRS SA-BERT+BERT-FP MAP 0.701 #2 of 7 Archive leaderboard report
Conversational Response Selection RRS SA-BERT+BERT-FP MRR 0.715 #2 of 7 Archive leaderboard report
Conversational Response Selection RRS SA-BERT+BERT-FP P@1 0.555 #2 of 7 Archive leaderboard report
Conversational Response Selection RRS SA-BERT+BERT-FP R10@1 0.497 #2 of 7 Archive leaderboard report
Conversational Response Selection RRS SA-BERT+BERT-FP R10@2 0.685 #2 of 7 Archive leaderboard report
Conversational Response Selection RRS SA-BERT+BERT-FP R10@5 0.931 #2 of 7 Archive leaderboard report
Conversational Response Selection RRS Ranking Test SA-BERT+BERT-FP NDCG@3 0.674 #2 of 4 Archive leaderboard report
Conversational Response Selection RRS Ranking Test SA-BERT+BERT-FP NDCG@5 0.753 #2 of 4 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) SA-BERT R10@1 0.855 #11 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) SA-BERT R10@2 0.928 #11 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) SA-BERT R10@5 0.983 #11 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu Dialogue (v1, Ranking) SA-BERT R2@1 0.965 #11 of 25 Archive leaderboard report
Conversational Response Selection Ubuntu IRC SA-BERT Accuracy 60.42 #2 of 5 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

AdamAttentionAttention DropoutBERTDense ConnectionsDropoutLayer NormalizationLinear LayerLinear Warmup With Linear DecayMulti-Head AttentionResidual ConnectionSoftmaxWeight DecayWordPiece

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