. . . . . . . . . . . . . . . . Language Generation via DAG Transduction Yajie Ye, Weiwei Sun and Xiaojun Wan {yeyajie,ws,wanxiaojun}@pku.edu.cn Institute of Computer Science and Technology Peking University . . . . . . . . . . . . . . . . . . . . . . . . July 17, 2018
. . . . . . . . . . . . . . . . . Overview 1 Background 2 Formal Models 3 Our DAG Transducer . . . . . . . . . . . . . . . . . . . . . . . 4 Evaluation
. . . . . . . . . . . . . . . . . Outline 1 Background 2 Formal Models 3 Our DAG Transducer . . . . . . . . . . . . . . . . . . . . . . . 4 Evaluation
. . . . . . . . . . . . A NLG system Architecture . Communicative Goal Document Planning Document Plans Microplanning Sentence Plans Linguistic Realisation Surface Text Reference Ehud Reiter and Robert Dale, Building Natural Language . . . . . . . . . . . . . . . . . . . . . . . . . . . Generation Systems, Cambridge University Press, 2000.
. . . . . . . . . . . . A NLG system Architecture . Communicative Goal Document Planning Document Plans Microplanning Sentence Plans Linguistic Realisation Surface Text In this paper, we study surface realization, i.e. mapping meaning . . . . . . . . . . . . . . . . . . . . . . . . . . . representations to natural language sentences.
. . . . . . . . . . . . . . . . . . Meaning Representation Feature structures . . . . . . . . . . . . . . . . . . . . . . This paper: Graphs! • Logic form, e.g. lambda calculus
. . . . . . . . . . . . . . . . . . Meaning Representation . . . . . . . . . . . . . . . . . . . . . . This paper: Graphs! • Logic form, e.g. lambda calculus • Feature structures
. . . . . . . . . . . . . . . . . . Meaning Representation . . . . . . . . . . . . . . . . . . . . . . • Logic form, e.g. lambda calculus • Feature structures • This paper: Graphs!
. . . . . . . . . . . . . . . . . . Graph-Structured Meaning Representation Difgerent kinds of graph-structured semantic representations: . . . . . . . . . . . . . . . . . . . . . . • Semantic Dependency Graphs (SDP) • Abstract Meaning Representations (AMR) • Dependency-based Minimal Recursion Semantics (DMRS) • Elementary Dependency Structures (EDS)
. . . . . . . . . . . . . . . . . . Graph-Structured Meaning Representation Difgerent kinds of graph-structured semantic representations: . . . . . . . . . . . . . . . . . . . . . . • Semantic Dependency Graphs (SDP) • Abstract Meaning Representations (AMR) • Dependency-based Minimal Recursion Semantics (DMRS) • Elementary Dependency Structures (EDS)
. . . . . . . . . . . . . . . . . . Graph-Structured Meaning Representation Difgerent kinds of graph-structured semantic representations: . . . . . . . . . . . . . . . . . . . . . . • Semantic Dependency Graphs (SDP) • Abstract Meaning Representations (AMR) • Dependency-based Minimal Recursion Semantics (DMRS) • Elementary Dependency Structures (EDS)
. _want_v_1 . . . . . . . . . Graph-Structured Meaning Representation Difgerent kinds of graph-structured semantic representations: _the_q . _boy_n_1 _the_q _believe_v_1 pronoun_q _girl_n_1 pron BV BV ARG1 ARG2 ARG1 ARG2 . . . . . . . . . . . . . . . . . BV . . . . . . . . . . . . • Semantic Dependency Graphs (SDP) • Abstract Meaning Representations (AMR) • Dependency-based Minimal Recursion Semantics (DMRS) • Elementary Dependency Structures (EDS)
. _the_q . . . . . . . . Type-Logical Semantic Graph EDS graphs are grounded under type-logical semantics. They are _want_v_1 _boy_n_1 . _the_q _believe_v_1 pronoun_q _girl_n_1 pron BV BV ARG1 ARG2 ARG1 ARG2 BV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The boy wants the girl to believe him. usually very fmat and multi-rooted graphs.
. . . . . . . . . . . . . . . Previous Work 1 Seqence-to-seqence Models. (AMR-to-text) 2 Synchronous Node Replacement Grammar. (AMR-to-text) 3 Other Unifjcation grammar-based methods Reference Ioannis Konstas, Srinivasan Iyer, Mark Yatskar, Yejin Choi, and Luke Zettlemoyer. 2017. Neural AMR: Sequence-to-sequence . . . . . . . . . . . . . . . . . . . . . . . . . models for parsing and generation.
. . . . . . . . . . . . . . . Previous Work 1 Seqence-to-seqence Models. (AMR-to-text) 2 Synchronous Node Replacement Grammar. (AMR-to-text) 3 Other Unifjcation grammar-based methods Reference Linfeng Song, Xiaochang Peng, Yue Zhang, Zhiguo Wang, and Daniel Gildea. 2017. AMR-to-text generation with synchronous . . . . . . . . . . . . . . . . . . . . . . . . . node replacement grammar.
. . . . . . . . . . . . . . . . Previous Work 1 Seqence-to-seqence Models. (AMR-to-text) 2 Synchronous Node Replacement Grammar. (AMR-to-text) 3 Other Unifjcation grammar-based methods Reference Carroll, John and Oepen, Stephan 2005. High effjciency realization . . . . . . . . . . . . . . . . . . . . . . . . for a wide-coverage unifjcation grammar
. . . . . . . . . . . . . . . . . Outline 1 Background 2 Formal Models 3 Our DAG Transducer . . . . . . . . . . . . . . . . . . . . . . . 4 Evaluation
. - . . . . . . . Formalisms for Strings, Trees and Graphs Chomsky hierarchy Grammar Abstract machines Type-0 Turing machine . Type-1 Context-sensitive Linear-bounded - Tree-adjoining Embedded pushdown Type-2 Context-free Nondeterministic pushdown Type-3 Regular Finite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manipulating Graphs: Graph Grammar and DAG Automata .
. . . . . . . . . . . . . . . . . . Existing System David Chiang, Frank Drewes, Daniel Gildea, Adam Lopez and . . . . . . . . . . . . . . . . . . . . . . the longest NLP paper that I’ve ever read Giorgio Satta. Weighted DAG Automata for Semantic Graphs.
. . . . . . . . . . . . . . DAG Automata A weighted DAG automaton is a tuple q 1 q 2 q 3 q m r 1 r 2 r n . . . . . . . . . . . . . . . . . . . . . . . . . . M = ⟨ Σ , Q , δ, K ⟩ . . . . . . ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ω σ σ ◦ ◦ . . . ◦ ◦ ◦ . . . ◦ σ/ω { q 1 , · · · , q m } − − → { r 1 , · · · , r n }
. . . . . . . . . . . . . . DAG Automata transitions: . q 1 r n q 2 r 2 q 3 r 1 . . q m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ω σ σ ◦ ◦ . . . ◦ ◦ ◦ . . . ◦ • A run of M on DAG D = ⟨ V , E , ℓ ⟩ is an edge labeling function ρ : E → Q . • The weight of ρ is the product of all weight of local [ ] ℓ ( v ) ⊗ δ ( ρ ) = δ ρ ( in ( v )) − − → ρ ( out ( v )) v ∈ V
. . . . . . . . . . . . DAG Automata: Toy Example . States: John wants to go. _want_v_1 _go_v_1 proper_q named(John) Recognition Rules: . _ want _ v _ 1 _ go _ v _ 1 . . proper _ q . . . . . . . . . . . . . . . . . . . . . . . . _ go _ v _ 1 {} − − − − − − − → { , } {} − − − − − − → { } { } − − − − − → { } { } − − − − − → { } named ( John ) { , , } − − − − − − − → {}
. . . . . . . . . . . . DAG Automata: Toy Example . States: John wants to go. _want_v_1 _go_v_1 proper_q named(John) Recognition Rules: . _ want _ v _ 1 _ go _ v _ 1 . . proper _ q . . . . . . . . . . . . . . . . . . . . . . . . _ go _ v _ 1 {} − − − − − − − → { , } {} − − − − − − → { } { } − − − − − → { } { } − − − − − → { } named ( John ) { , , } − − − − − − − → {}
. . . . . . . . . . . . DAG Automata: Toy Example . States: John wants to go. _want_v_1 _go_v_1 proper_q named(John) Recognition Rules: . _ want _ v _ 1 _ go _ v _ 1 . . proper _ q . . . . . . . . . . . . . . . . . . . . . . . . _ go _ v _ 1 {} − − − − − − − → { , } {} − − − − − − → { } { } − − − − − → { } { } − − − − − → { } named ( John ) { , , } − − − − − − − → {}
. . . . . . . . . . . . DAG Automata: Toy Example . States: John wants to go. _want_v_1 _go_v_1 proper_q named(John) Recognition Rules: . _ want _ v _ 1 _ go _ v _ 1 . . proper _ q . . . . . . . . . . . . . . . . . . . . . . . . _ go _ v _ 1 {} − − − − − − − → { , } {} − − − − − − → { } { } − − − − − → { } { } − − − − − → { } named ( John ) { , , } − − − − − − − → {}
Recommend
More recommend