Filters








189 Hits in 7.3 sec

Weakly-Supervised Neural Semantic Parsing with a Generative Ranker

Jianpeng Cheng, Mirella Lapata
2018 Proceedings of the 22nd Conference on Computational Natural Language Learning  
Experiments on three Freebase datasets demonstrate the effectiveness of our semantic parser, achieving state-of-the-art results.  ...  ., whether they are likely to execute to the correct denotation and the degree to which they preserve the meaning of the utterance; (b) a scheduled training procedure effectively balances the contribution  ...  Acknowledgments This research is supported by a Google PhD Fellowship and an AdeptMind Scolar Fellowship to the first author.  ... 
doi:10.18653/v1/k18-1035 dblp:conf/conll/0001L18 fatcat:dr7ox2u5b5agdfawmrrji2d4da

Weakly-supervised Neural Semantic Parsing with a Generative Ranker [article]

Jianpeng Cheng, Mirella Lapata
2018 arXiv   pre-print
Experiments on three Freebase datasets demonstrate the effectiveness of our semantic parser, achieving results within the state-of-the-art range.  ...  In this paper we introduce a neural parser-ranker system for weakly-supervised semantic parsing.  ...  Acknowledgments This research is supported by a Google PhD Fellowship and an AdeptMind Scolar Fellowship to the first author.  ... 
arXiv:1808.07625v1 fatcat:6tg5z4rjfzbijfj7mcbxwheccu

Machine Reading Comprehension Using Structural Knowledge Graph-aware Network

Delai Qiu, Yuanzhe Zhang, Xinwei Feng, Xiangwen Liao, Wenbin Jiang, Yajuan Lyu, Kang Liu, Jun Zhao
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
Leveraging external knowledge is an emerging trend in machine comprehension task.  ...  To this end, we propose a Structural Knowledge Graph-aware Network (SKG) model, constructing sub-graphs for entities in the machine comprehension context.  ...  Acknowledgements This work was supported by National Key R&D Program of China under Grant 2018YFB1005100, the National Natural Science Foundation of China (No.61533018, No.U1605251).  ... 
doi:10.18653/v1/d19-1602 dblp:conf/emnlp/QiuZFLJLLZ19 fatcat:ctjguxojo5h55f7mulg6zrpxb4

Discrete-State Variational Autoencoders for Joint Discovery and Factorization of Relations

Diego Marcheggiani, Ivan Titov
2016 Transactions of the Association for Computational Linguistics  
We present a method for unsupervised open-domain relation discovery.  ...  The model is composed of two parts: a feature-rich relation extractor, which predicts a semantic relation between two entities, and a factorization model, which reconstructs arguments (i.e., the entities  ...  The authors thank 5 github.com/diegma/relation-autoencoder 242 the action editor and the anonymous reviewers for their valuable suggestions and Limin Yao for answering our questions about data and baselines  ... 
doi:10.1162/tacl_a_00095 fatcat:xwr5n3h23navrejh4q72ep4gzi

Recurrent One-Hop Predictions for Reasoning over Knowledge Graphs [article]

Wenpeng Yin, Yadollah Yaghoobzadeh, Hinrich Schütze
2018 arXiv   pre-print
Reasoning over multi-hop (mh) KG paths is thus an important capability that is needed for question answering or other NLP tasks that require knowledge about the world. mh-KG reasoning includes diverse  ...  Our models show state-of-the-art for two important multi-hop KG reasoning tasks: Knowledge Base Completion and Path Query Answering.  ...  Acknowledgement We gratefully acknowledge the support of the European Research Council for this research (ERC Advanced Grant NonSequeToR, # 740516).  ... 
arXiv:1806.04523v1 fatcat:5nprpxdqprh6pp44s6lv2n5q7m

Measuring Compositional Generalization: A Comprehensive Method on Realistic Data [article]

Daniel Keysers, Nathanael Schärli, Nathan Scales, Hylke Buisman, Daniel Furrer, Sergii Kashubin, Nikola Momchev, Danila Sinopalnikov, Lukasz Stafiniak, Tibor Tihon, Dmitry Tsarkov, Xiao Wang (+2 others)
2020 arXiv   pre-print
We present a large and realistic natural language question answering dataset that is constructed according to this method, and we use it to analyze the compositional generalization ability of three machine  ...  systematically construct such benchmarks by maximizing compound divergence while guaranteeing a small atom divergence between train and test sets, and we quantitatively compare this method to other approaches for  ...  CFQ is a simple yet realistic, large dataset of natural language questions and answers that also provides for each question a corresponding SPARQL query against the Freebase knowledge base (Bollacker  ... 
arXiv:1912.09713v2 fatcat:qpqftadxqbaengbo3dq726dgem

Structured Knowledge Discovery from Massive Text Corpus [article]

Chenwei Zhang
2019 arXiv   pre-print
In this dissertation, I will introduce principles, models, and algorithms for effective structured knowledge discovery from the massive text corpus.  ...  In particular, four problems are studied in this dissertation: Structured Intent Detection for Natural Language Understanding, Structure-aware Natural Language Modeling, Generative Structured Knowledge  ...  For example, online question answering websites are able to offer globally accessible information via human-human interactions.  ... 
arXiv:1908.01837v1 fatcat:j46srlxblfd35cd4z6jkl43iiu

Semantic Parsing with Dual Learning

Ruisheng Cao, Su Zhu, Chen Liu, Jieyu Li, Kai Yu
2019 Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics  
Semantic parsing converts natural language queries into structured logical forms. The paucity of annotated training samples is a fundamental challenge in this field.  ...  In this work, we develop a semantic parsing framework with the dual learning algorithm, which enables a semantic parser to make full use of data (labeled and even unlabeled) through a dual-learning game  ...  Learning to collaborate for question answering and asking.  ... 
doi:10.18653/v1/p19-1007 dblp:conf/acl/CaoZLLY19 fatcat:tbsrr24ij5exphjctkmwwwf4l4

Question Answering over Knowledge Graphs via Structural Query Patterns [article]

Weiguo Zheng, Mei Zhang
2019 arXiv   pre-print
Natural language question answering over knowledge graphs is an important and interesting task as it enables common users to gain accurate answers in an easy and intuitive manner.  ...  Searching the structured query over the knowledge graph can produce answers to the question.  ...  In this paper, we focus on constructing query graphs for answering natural language questions over a knowledge graph.  ... 
arXiv:1910.09760v2 fatcat:oghjct4335grvnyuhhk4tovdxy

GraphQ IR: Unifying Semantic Parsing of Graph Query Language with Intermediate Representation [article]

Lunyiu Nie, Shulin Cao, Jiaxin Shi, Qi Tian, Lei Hou, Juanzi Li, Jidong Zhai
2022 arXiv   pre-print
With the IR's natural-language-like representation that bridges the semantic gap and its formally defined syntax that maintains the graph structure, neural semantic parser can more effectively convert  ...  In this paper, we propose a unified intermediate representation (IR) for graph query languages, namely GraphQ IR.  ...  ., 2020b ) is a large-scale dataset for complex question answering over Wikidata knowledge base (Vrandečić and Krötzsch, 2014) .  ... 
arXiv:2205.12078v1 fatcat:ewv6iyoyvjfepko2cbw4v7il5m

Towards End-to-End Learning for Dialog State Tracking and Management using Deep Reinforcement Learning

Tiancheng Zhao, Maxine Eskenazi
2016 Proceedings of the 17th Annual Meeting of the Special Interest Group on Discourse and Dialogue  
This paper presents an end-to-end framework for task-oriented dialog systems using a variant of Deep Recurrent Q-Networks (DRQN).  ...  We evaluated the proposed model on a 20 Question Game conversational game simulator.  ...  Also there is a 5% chance that the simulator will consider unknown as an attribute and thus it will answer with unknown intent for any question related to it.  ... 
doi:10.18653/v1/w16-3601 dblp:conf/sigdial/ZhaoE16 fatcat:aundy2dhmjdldndc2fveom7foy

Towards End-to-End Learning for Dialog State Tracking and Management using Deep Reinforcement Learning [article]

Tiancheng Zhao, Maxine Eskenazi
2016 arXiv   pre-print
This paper presents an end-to-end framework for task-oriented dialog systems using a variant of Deep Recurrent Q-Networks (DRQN).  ...  We evaluated the proposed model on a 20 Question Game conversational game simulator.  ...  We would also like to thank Alan W Black for discussions on this paper.  ... 
arXiv:1606.02560v2 fatcat:i4akopezyzchzbonjfx57pecdi

Neural Approaches to Conversational AI [article]

Jianfeng Gao, Michel Galley, Lihong Li
2019 arXiv   pre-print
We group conversational systems into three categories: (1) question answering agents, (2) task-oriented dialogue agents, and (3) chatbots.  ...  For each category, we present a review of state-of-the-art neural approaches, draw the connection between them and traditional approaches, and discuss the progress that has been made and challenges still  ...  Figure 6 . 3 : 63 Given a complex question Q, we decompose it to a sequence of simple questions Q 1 , Q 2 , ..., use a Web-scale KB-QA agent to generate for each Q i an answer A i , from which we compute  ... 
arXiv:1809.08267v3 fatcat:j57xlm4ogferdnrpfs4f2jporq

Benchmarking Knowledge Graphs on the Web [article]

Michael Röder and Mohamed Ahmed Sherif and Muhammad Saleem and Felix Conrads and Axel-Cyrille Ngonga Ngomo
2020 arXiv   pre-print
The growing interest in making use of Knowledge Graphs for developing explainable artificial intelligence, there is an increasing need for a comparable and repeatable comparison of the performance of Knowledge  ...  This paper gives an overview of benchmarks used to evaluate systems that process Knowledge Graphs.  ...  Matching URIs and labels can be a particularly difficult task for a question answering benchmarking framework. • C2KB: This sub-task aims to identify all relevant resources for the given question.  ... 
arXiv:2002.06039v1 fatcat:tc25o6rplbekndjng4cqh2oxoe

The VQA-Machine: Learning How to Use Existing Vision Algorithms to Answer New Questions

Peng Wang, Qi Wu, Chunhua Shen, Anton van den Hengel
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
One of the most intriguing features of the Visual Question Answering (VQA) challenge is the unpredictability of the questions.  ...  Extracting the information required to answer them demands a variety of image operations from detection and counting, to segmentation and reconstruction.  ...  [30] propose a VQA framework named "Ahab" that uses explicit reasoning over an RDF (Resource Description Framework) Knowledge Base to derive the answer, which naturally gives rise to a reasoning chain  ... 
doi:10.1109/cvpr.2017.416 dblp:conf/cvpr/WangWSH17 fatcat:m7f32lzpsrd73iys2xawv7rr7e
« Previous Showing results 1 — 15 out of 189 results