Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision

By Chen Liang, Jonathan Berant, Quoc V. Le, Kenneth D. Forbus, Ni Lao. 2016

427 citations recorded by the source corpus.

Harnessing the statistical power of neural networks to perform language understanding and symbolic reasoning is difficult, when it requires executing efficient discrete operations against a large knowledge-base. In this work, we introduce a Neural Symbolic Machine, which contains (a) a neural “programmer”, i.e., a sequence-to-sequence model that maps language utterances to programs and utilizes a key-variable memory to handle compositionality (b) a symbolic “computer”, i.e., a Lisp interpreter that performs program execution, and helps find good programs by pruning the search space. We apply REINFORCE to directly optimize the task reward of this structured prediction problem. To train with weak supervision and improve the stability of REINFORCE, we augment it with an iterative maximum-likelihood training process. NSM outperforms the state-of-the-art on the WebQuestionsSP dataset when trained from question-answer pairs only, without requiring any feature engineering or domain-specific knowledge.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Natural Language Processing

Citations in the corpus

DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs (2019, background): “Other re-lated work along these lines has been done by Reed and de Freitas (2015), Neelakantan et al. (2015), and Liang et al. (2017).”
Program Induction by Rationale Generation: Learning to Solve and Explain Algebraic Word Problems (2017, background): “This is analogous to the question answering work in Liang et al. (2016), where the query that generates the correct answer must be found during inference, and training proved to be difﬁcult without supervision.”
A Discrete Hard EM Approach for Weakly Supervised Question Answering (2019, methodology): “…QAN ET (Yu et al., 2018a) which selects two numeric values from the document or the question and performs addition or subtraction to get the an-6 (Liang et al., 2017) 70.1 Hard EM (Liang et al., 2018) 70.2 Beam-based MML (Liang et al., 2018) 70.7 MAPO (Liang et al., 2018) 71.8 72.4 MAPOX…”
Improving a Neural Semantic Parser by Counterfactual Learning from Human Bandit Feedback (2018, background): “More recently, several works have applied policy gradient techniques such as REINFORCE (Williams, 1992) to train neural semantic parsers (Liang et al. (2017); Mou et al. (2017); Peng et al. (2017); inter alia ).”
Reliability and Learnability of Human Bandit Feedback for Sequence-to-Sequence Reinforcement Learning (2018, methodology): “A typical approach is to combine REINFORCE (Williams, 1992) with policies based on deep sequence-to-sequence learning (Bahdanau et al., 2015), for example, in machine translation (Bahdanau et al., 2017), semantic parsing (Liang et al., 2017), or summarization (Paulus et al., 2017).”
A Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning (2019, methodology): “Later works on program induction (Reed and De Freitas, 2016; Neelakantan et al., 2016; Liang et al., 2017) extended this idea by using several built-in logic operations along with a key- value memory to learn different types of compositional programs such as addition or sorting.”
Inferring and Executing Programs for Visual Reasoning (2017, background): “Recent approaches to semantic parsing involve a learnt programmer [28].”
AutoML-Zero: Evolving Machine Learning Algorithms From Scratch (2020, background): “More recently, program synthesis has focused on solving problems like sorting (Graves et al., 2014), string manipulation (Gulwani et al., 2017; Balog et al., 2017), or structured data QA (Liang et al., 2016).”
Self-Imitation Learning (2018, methodology): “Although a similar idea has been discussed for program synthesis (Liang et al., 2016; Abolafia et al., 2018), this prior work used classification loss without justification.”
Troubling Trends in Machine Learning Scholarship (2018, background): “t learning [30] and generative adversarial networks [51]. See [68] for more discussion of lapses in empirical rigor and resulting consequences. In contrast, many papers perform good ablation analyses [41, 45, 77, 82], and even retrospective attempts to isolate the source of gains can lead to new dis”
Search-based Neural Structured Learning for Sequential Question Answering (2017, background): “Other than FP and NP , the work of Neural Symbolic Machines (NSM) (Liang et al., 2017) is perhaps the closest to ours.”
Ask the Right Questions: Active Question Reformulation with Reinforcement Learning (2017, background): “Uses of policy gradient for QA include Liang et al. [2017], who train a semantic parser to query a knowledge base, and Seo et al. [2017b] who propose query reduction networks that transform a query to answer questions that involve multi-hop common sense reasoning.”
TabFact: A Large-scale Dataset for Table-based Fact Verification (2019, methodology): “Thus, it can be seen as a weakly supervised learning problem as discussed in Liang et al. (2017); Lao et al. (2011).”
Improved Neural Relation Detection for Knowledge Base Question Answering (2017, methodology): “For example, our model could be integrated into the de-coder in (Liang et al., 2016), to provide better sequence prediction.”
From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood (2017, methodology): “, 2016), and even semantic parsing (Liang et al., 2017).”
Coupling Retrieval and Meta-Learning for Context-Dependent Semantic Parsing (2019, background): “…logic (Zelle and Mooney, 1995; Clarke et al., 2010), lambda calculus (Zettlemoyer and Collins, 2005; Kushman and Barzilay, 2013), dependency-based compositional semantics (Liang et al., 2011; Berant et al., 2013), and database queries (Iyer et al., 2017; Zhong et al., 2017; Liang et al., 2017).”
Neural Program Synthesis with Priority Queue Training (2018, methodology): “, ﬁlling missing values in tables (Devlin et al., 2017), and querying tables (Liang et al., 2017).”