Making Neural QA as Simple as Possible but not Simpler
Explore this paper's citation graph
Summary
This work proposes a simple heuristic that guides the development of neural baseline systems for the extractive QA task and finds that there are two ingredients necessary for building a high-performing neural QA system: the awareness of question words while processing the context and a composition function that goes beyond simple bag-of-words modeling, such as recurrent neural networks.
- Type
- article
- Published
- 2017-03-14
- Cited by
- 220
- References
- 29
- Access
- Open access
- OpenAlex
- https://openalex.org/W2626154462
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:2592133
Keywords
Computer science, Simple (philosophy), Artificial intelligence, Task (project management), Artificial neural network
References
- Introduction to "This is Watson"
- Teaching Machines to Read and Comprehend
- LASSO: A Tool for Surfing the Answer Net
- Long Short-Term Memory
- MCTest: A Challenge Dataset for the Open-Domain Machine Comprehension of Text
- The Goldilocks Principle: Reading Children's Books with Explicit Memory Representations
- Automated question answering: review of the main approaches
- GloVe: Global Vectors for Word Representation
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- Neural Architectures for Named Entity Recognition
- Dataset and Neural Recurrent Sequence Labeling Model for Open-Domain Factoid Question Answering
- Machine Comprehension Using Match-LSTM and Answer Pointer
- ReasoNet: Learning to Stop Reading in Machine Comprehension
- End-to-End Reading Comprehension with Dynamic Answer Chunk Ranking
- NewsQA: A Machine Comprehension Dataset
- MS MARCO: A Human Generated MAchine Reading COmprehension Dataset
- Multi-Perspective Context Matching for Machine Comprehension
- Overview of the TREC 2007 Question Answering Track
- Reading Wikipedia to Answer Open-Domain Questions
- Bidirectional Attention Flow for Machine Comprehension
Cited by
- Ruminating Reader: Reasoning with Gated Multi-hop Attention
- Reading Twice for Natural Language Understanding
- Neural Domain Adaptation for Biomedical Question Answering
- Neural Question Answering at BioASQ 5B
- Reinforced Mnemonic Reader for Machine Reading Comprehension
- Globally Normalized Reader
- Constructing Datasets for Multi-hop Reading Comprehension Across Documents
- Simple and Effective Multi-Paragraph Reading Comprehension
- DCN+: Mixed Objective and Deep Residual Coattention for Question Answering
- Dynamic Integration of Background Knowledge in Neural NLU Systems
- Neural Skill Transfer from Supervised Language Tasks to Reading Comprehension
- Dynamic Fusion Networks for Machine Reading Comprehension
- Stochastic Answer Networks for Machine Reading Comprehension
- Contextualized Word Representations for Reading Comprehension
- A Question-Focused Multi-Factor Attention Network for Question Answering
- DuoRC: Towards Complex Language Understanding with Paraphrased Reading Comprehension
- A Study of Word Embeddings for Biomedical Question Answering
- Translating Questions into Answers using DBPedia n-triples
- Know What You Don’t Know: Unanswerable Questions for SQuAD
- Verification of the Expected Answer Type for Biomedical Question Answering
Related papers
- Overview of Question-Answering
- Study and application of telecom operators system security state baseline
- Analysis of the Baseline Decorrelation and Critical Baseline of Interferometric SAR
- Research of Baseline Implement
- Accounting for baseline targets in NDCs: Issues and options for guidance
- A Survey on Question and Answering Systems
- Study on Baseline Evaluation Application in Science and Technology Assessment
- The Baseline: A Social Construction
- Baseline choice and performance implications for REDD