路径选择是知识库问答任务的关键步骤,语义相似度常被用来计算路径对于问句的相似度得分。针对测试集中存在大量未见的关系,该文提出使用一种负例动态采样的语义相似度模型的训练方法,去丰富训练集中关系的多样性,模型性能得到显著提升。针对复杂问题候选路径数量组合爆炸问题,该文比较了两种路径剪枝方法,即基于分类的方法和基于集束搜索的方法。在包含简单问题和复杂问题的CCKS 2019-CKBQA评测数据集上,该方法能达到较优异的性能,测试集上单模型系统平均F1值达到0.694,系统融合后达到0.731。
Abstract
Path selection, as a key step in the Knowledge Base Question Answering (KBQA) task, relies on the the semantic similarity between a question and candidate paths. To deal with massive unseen relations in the test set, a method based on dynamic sampling of negative examples is proposed to enrich the relations in the training set. In the prediction phase, two path pruning methods, i.e., the classification method and the beam search method, are compared to tackle the explosion of candidate paths. On the CCKS 2019-CKBQA evaluation data set containing simple and complex problems, the proposed method achieves an average F1 value of 0.694 for the single-model system, and 0.731 for the ensemble system.
关键词
知识库问答 /
BERT /
动态采样 /
集束搜索
{{custom_keyword}} /
Key words
KBQA /
BERT /
dynamic sampling /
beam search
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Bollacker K, Cook R, Tufts P. Freebase: A shared database of structured general human knowledge[C]//Proceedings of the 22nd National Conference on Artificial Intelligence, 2007: 1962-1963.
[2] Bizer C, Lehmann J, Kobilarov G, et al. DBpedia: A crystallization point for the web of data[J]. Web Semantics Science Services and Agents on the World Wide Web, 2009, 7(3): 154-165.
[3] Niu X, Sun X, Wang H, et al. Zhishi.me: Weaving Chinese linking open data[C]//Proceedings of the Semantic Web - ISWC 2011. Lecture Notes in Computer Science, 2011: 205-220.
[4] Xu B, Xu Y, Liang J Q, et al. CNDB pedia: A never-ending Chinese knowledge extraction system[C]//Proceedings of the International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, 2017: 428-438.
[5] Bao J W, Duan N, Yan Z, et al. Constraint-based question answering with knowledge graph[C]//Proceedings of the 26th International Conference on Computational Linguistics: Technical Papers, 2016: 2503-2514.
[6] Bordes A, Chopra S, Weston J. Question answering with subgraph embeddings[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2014: 615-620.
[7] Berant J, Chou A, Frostig R, et al. Semantic parsing on freebase from question-answer pairs[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2013: 1533-1544.
[8] Reddy S, Lapata M, Steedman M. Large-scale semantic parsing without question answer pairs[C]//Proceedings of the Transactions of the Association for Computational Linguistics, 2014: 377-392.
[9] Yin W, Chang M W, He X D, et al. Semantic parsing via staged query graph generation: Question answering with knowledge base[C]//Proceedings of the International Joint Conference on Natural Language Processing, 2015: 1321-1331.
[10] Mohammed S, Shi P, Lin J. Strong baselines for simple question answering over knowledge graphs with and without neural networks[C]//Proceedings of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2018: 291-296.
[11] Yu M, Yin W P, Hasan K S, et al. Improved neural relation detection for knowledge base question answering[C]//Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, 2017: 571-581.
[12] Wang Y, Zhang R C, Xu C, et al, The APVA-TURBO approach to question answering in knowledge base[C]//Proceedings of the 27th International Conference on Computational Linguistics, 2018: 1998-2009.
[13] Xu K, Reddy S, Feng Y S, et al. Question answering on freebase via relation extraction and textual evidence[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, 2016: 2326-2336.
[14] Petrochuk M, Zettlemoyer L. Simple questions nearly solved: A new upperbound and baseline approach[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2018: 554-558.
[15] Chen Z Y, Chang C, Chen Y P, et al. U-Hop: An unrestricted-hop relation extraction framework for knowledge-based question answering[C]//Proceedings of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2019: 345-356.
[16] Katiyar A, Cardie C. Investigating LSTMS for joint extraction of opinion entities and relations[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, 2016: 919-929.
[17] Luo K Q, Lin F L, Luo X S, et al. Knowledge base question answering via encoding of complex query graphs[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2018: 2185-2194.
[18] Peters M E, Neumann M, Iyyer M, et al. Deep contextualized word representations[C]//Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2018: 2227-2237.
[19] Radford A, Narasimhan K, Salimans T, et al. Improving language understanding by generative pre-training[EB/OL].https://s3-us-west-2.amazonaws.com/openaiassets/research-covers/languageunsupervised/language understanding paper.Pdf.[2020-12-10].
[20] Devlin J, Chang M W, LEE K, et al. BERT: Pre-training of deep bidirectional transformers for language understanding[C]//Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2018: 4171-4186.
[21] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//Proceedings of the Advances in Neural Information Processing Systems, 2017: 6000-6010.
[22] Garg S, VU T, Moschitti A. Tanda: Transfer and adapt pre-trained transformer models for answer sentence selection[EB/OL]. https://arxiv.org/abs/1911.04118. [2020-12-10].
[23] Zhang P J, Wu K, Zhu Z K, et al. Combining neural network models with rules for Chinese knowledge base question answering[EB/OL]. https://conference.bj.bcebos.com/ccks2019/eval/webpage/index.html. [2020-12-10].
[24] 骆金昌,尹存祥,吴晓晖,等. 混合语义相似度的中文知识图谱问答系统[EB/OL]. https://conference.bj.bcebos.com/ccks2019/eval/webpage/index.html.[2020-12-10].
[25] Yang Y Y, He X H, Zhou K J, et al. Multi-module system for open domain Chinese question answering over knowledge base[EB/OL]. https://conference.bj.bcebos.com/ccks2019/eval/webpage/index.html. [2020-12-10].
[26] 曹明宇,李帅驰,王鑫雷,等. DUTIR中文开放域知识库问答评测报告[EB/OL]. https://conference.bj.bcebos.com/ccks2019/eval/webpage/index.html. [2019-08-26].
[27] Yin W P, Yu M, Xiang B, et al. Simple question answering by attentive convolutional neural network[C]//Proceedings of the 26th International Conference on Computational Linguistics, Proceedings of the Conference, 2016: 1746-1756.
[28] Kingma D P, Ba J L. Adam: A method for stochastic optimiztion[EB/OL]. https://arxiv.org/abs/1412.6980v8. [2020-12-10].
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金(61702518,61876116)
{{custom_fund}}
PDF(3129 KB)
