How to Generate Reasonable Texts with Controlled Attributes View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2020-09-22

AUTHORS

Yanan Zheng , Yan Wang , Lijie Wen , Jianmin Wang

ABSTRACT

The controllable text generation (CTG) task is crucial for text-related applications, such as goal-oriented dialogue systems and text style-transfer applications, etc. However, existing CTG methods commonly ignore the co-occurrence dependencies between multiple controlled attributes, which are implicit in domain knowledge. As a result, rarely co-occurring controlled values are highly likely to be given by users, which finally leads to non-committal generated texts that are out of control. To address this problem, we initially propose the Dependency-aware Controllable Text Generation (DCTG) model that reduces trivial generations by automatically learning the co-occurrence dependencies and adjusting rarely co-occurring controlled values. Our DCTG highlights in (1) modeling the co-occurrence dependencies between controlled attributes with neural networks, (2) integrating dependency losses to guide each component of our model to collaboratively work for generating reasonable texts based on the learned dependencies, and (3) proposing a novel Reasonableness metric measuring to which degree generations comply with real co-occurrence dependencies. Experiments prove that DCTG outperforms state-of-the-art baselines on three datasets in multiple aspects. More... »

PAGES

245-262

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-030-59416-9_15

DOI

http://dx.doi.org/10.1007/978-3-030-59416-9_15

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1131075133


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Information and Computing Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "School of Software, Tsinghua University, Beijing, China", 
          "id": "http://www.grid.ac/institutes/grid.12527.33", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Zheng", 
        "givenName": "Yanan", 
        "id": "sg:person.010412660037.85", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010412660037.85"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "School of Information, Central University of Finance and Economics, Beijing, China", 
          "id": "http://www.grid.ac/institutes/grid.411054.5", 
          "name": [
            "School of Information, Central University of Finance and Economics, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Wang", 
        "givenName": "Yan", 
        "id": "sg:person.012005621037.94", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012005621037.94"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "School of Software, Tsinghua University, Beijing, China", 
          "id": "http://www.grid.ac/institutes/grid.12527.33", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Wen", 
        "givenName": "Lijie", 
        "id": "sg:person.013640554311.55", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013640554311.55"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "School of Software, Tsinghua University, Beijing, China", 
          "id": "http://www.grid.ac/institutes/grid.12527.33", 
          "name": [
            "School of Software, Tsinghua University, Beijing, China"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Wang", 
        "givenName": "Jianmin", 
        "id": "sg:person.012303351315.43", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012303351315.43"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2020-09-22", 
    "datePublishedReg": "2020-09-22", 
    "description": "The controllable text generation (CTG) task is crucial for text-related applications, such as goal-oriented dialogue systems and text style-transfer applications, etc. However, existing CTG methods commonly ignore the co-occurrence dependencies between multiple controlled attributes, which are implicit in domain knowledge. As a result, rarely co-occurring controlled values are highly likely to be given by users, which finally leads to non-committal generated texts that are out of control. To address this problem, we initially propose the Dependency-aware Controllable Text Generation (DCTG) model that reduces trivial generations by automatically learning the co-occurrence dependencies and adjusting rarely co-occurring controlled values. Our DCTG highlights in (1) modeling the co-occurrence dependencies between controlled attributes with neural networks, (2) integrating dependency losses to guide each component of our model to collaboratively work for generating reasonable texts based on the learned dependencies, and (3) proposing a novel Reasonableness metric measuring to which degree generations comply with real co-occurrence dependencies. Experiments prove that DCTG outperforms state-of-the-art baselines on three datasets in multiple aspects.", 
    "editor": [
      {
        "familyName": "Nah", 
        "givenName": "Yunmook", 
        "type": "Person"
      }, 
      {
        "familyName": "Cui", 
        "givenName": "Bin", 
        "type": "Person"
      }, 
      {
        "familyName": "Lee", 
        "givenName": "Sang-Won", 
        "type": "Person"
      }, 
      {
        "familyName": "Yu", 
        "givenName": "Jeffrey Xu", 
        "type": "Person"
      }, 
      {
        "familyName": "Moon", 
        "givenName": "Yang-Sae", 
        "type": "Person"
      }, 
      {
        "familyName": "Whang", 
        "givenName": "Steven Euijong", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-030-59416-9_15", 
    "inLanguage": "en", 
    "isAccessibleForFree": false, 
    "isPartOf": {
      "isbn": [
        "978-3-030-59415-2", 
        "978-3-030-59416-9"
      ], 
      "name": "Database Systems for Advanced Applications", 
      "type": "Book"
    }, 
    "keywords": [
      "co-occurrence dependencies", 
      "goal-oriented dialogue systems", 
      "text-related applications", 
      "text generation tasks", 
      "text generation model", 
      "domain knowledge", 
      "art baselines", 
      "dialogue systems", 
      "neural network", 
      "generation task", 
      "generation model", 
      "DCTg", 
      "attributes", 
      "text", 
      "users", 
      "dependency", 
      "dataset", 
      "network", 
      "applications", 
      "task", 
      "multiple aspects", 
      "model", 
      "generation", 
      "system", 
      "knowledge", 
      "measuring", 
      "method", 
      "experiments", 
      "aspects", 
      "components", 
      "results", 
      "state", 
      "control", 
      "values", 
      "loss", 
      "baseline", 
      "problem", 
      "controllable text generation (CTG) task", 
      "text style-transfer applications", 
      "style-transfer applications", 
      "CTG methods", 
      "Dependency-aware Controllable Text Generation (DCTG) model", 
      "Controllable Text Generation (DCTG) model", 
      "trivial generations", 
      "dependency losses", 
      "reasonable texts", 
      "novel Reasonableness metric measuring", 
      "Reasonableness metric measuring", 
      "metric measuring", 
      "degree generations", 
      "real co-occurrence dependencies", 
      "Generate Reasonable Texts"
    ], 
    "name": "How to Generate Reasonable Texts with Controlled Attributes", 
    "pagination": "245-262", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1131075133"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-030-59416-9_15"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-030-59416-9_15", 
      "https://app.dimensions.ai/details/publication/pub.1131075133"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-01-01T19:23", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/chapter/chapter_39.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-030-59416-9_15"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-59416-9_15'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-59416-9_15'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-59416-9_15'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-59416-9_15'


 

This table displays all metadata directly associated to this object as RDF triples.

161 TRIPLES      23 PREDICATES      77 URIs      70 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-030-59416-9_15 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author Ncf3286ff278b4c0ea58aea12171d14c7
4 schema:datePublished 2020-09-22
5 schema:datePublishedReg 2020-09-22
6 schema:description The controllable text generation (CTG) task is crucial for text-related applications, such as goal-oriented dialogue systems and text style-transfer applications, etc. However, existing CTG methods commonly ignore the co-occurrence dependencies between multiple controlled attributes, which are implicit in domain knowledge. As a result, rarely co-occurring controlled values are highly likely to be given by users, which finally leads to non-committal generated texts that are out of control. To address this problem, we initially propose the Dependency-aware Controllable Text Generation (DCTG) model that reduces trivial generations by automatically learning the co-occurrence dependencies and adjusting rarely co-occurring controlled values. Our DCTG highlights in (1) modeling the co-occurrence dependencies between controlled attributes with neural networks, (2) integrating dependency losses to guide each component of our model to collaboratively work for generating reasonable texts based on the learned dependencies, and (3) proposing a novel Reasonableness metric measuring to which degree generations comply with real co-occurrence dependencies. Experiments prove that DCTG outperforms state-of-the-art baselines on three datasets in multiple aspects.
7 schema:editor N4e9a2d99d9ae46ee9d70a441152debca
8 schema:genre chapter
9 schema:inLanguage en
10 schema:isAccessibleForFree false
11 schema:isPartOf Ndeaffae5bdea4ebe852f58397c1f4df5
12 schema:keywords CTG methods
13 Controllable Text Generation (DCTG) model
14 DCTg
15 Dependency-aware Controllable Text Generation (DCTG) model
16 Generate Reasonable Texts
17 Reasonableness metric measuring
18 applications
19 art baselines
20 aspects
21 attributes
22 baseline
23 co-occurrence dependencies
24 components
25 control
26 controllable text generation (CTG) task
27 dataset
28 degree generations
29 dependency
30 dependency losses
31 dialogue systems
32 domain knowledge
33 experiments
34 generation
35 generation model
36 generation task
37 goal-oriented dialogue systems
38 knowledge
39 loss
40 measuring
41 method
42 metric measuring
43 model
44 multiple aspects
45 network
46 neural network
47 novel Reasonableness metric measuring
48 problem
49 real co-occurrence dependencies
50 reasonable texts
51 results
52 state
53 style-transfer applications
54 system
55 task
56 text
57 text generation model
58 text generation tasks
59 text style-transfer applications
60 text-related applications
61 trivial generations
62 users
63 values
64 schema:name How to Generate Reasonable Texts with Controlled Attributes
65 schema:pagination 245-262
66 schema:productId N4e71e6fd237349e081654ee731534fc8
67 Nc449e37db74b463992f12c94b12d4c77
68 schema:publisher Ne6ae7f9bab0a4d4b9c6ccc5cf5b3ef3c
69 schema:sameAs https://app.dimensions.ai/details/publication/pub.1131075133
70 https://doi.org/10.1007/978-3-030-59416-9_15
71 schema:sdDatePublished 2022-01-01T19:23
72 schema:sdLicense https://scigraph.springernature.com/explorer/license/
73 schema:sdPublisher N408dadedac5c40708e5a3e6784ab36df
74 schema:url https://doi.org/10.1007/978-3-030-59416-9_15
75 sgo:license sg:explorer/license/
76 sgo:sdDataset chapters
77 rdf:type schema:Chapter
78 N1ac7bf1399c6422e8c2682e158c4cfd3 schema:familyName Cui
79 schema:givenName Bin
80 rdf:type schema:Person
81 N357540b5128940ca82c8cb139e72cae2 schema:familyName Nah
82 schema:givenName Yunmook
83 rdf:type schema:Person
84 N408dadedac5c40708e5a3e6784ab36df schema:name Springer Nature - SN SciGraph project
85 rdf:type schema:Organization
86 N4d6e50e8d40447249b620c4ef4020d50 rdf:first sg:person.012303351315.43
87 rdf:rest rdf:nil
88 N4e71e6fd237349e081654ee731534fc8 schema:name doi
89 schema:value 10.1007/978-3-030-59416-9_15
90 rdf:type schema:PropertyValue
91 N4e9a2d99d9ae46ee9d70a441152debca rdf:first N357540b5128940ca82c8cb139e72cae2
92 rdf:rest N9d9cb75ba0e6423e9edd4ce78498707b
93 N5ef7fdc03e5c4256ae07b6a5883da885 schema:familyName Moon
94 schema:givenName Yang-Sae
95 rdf:type schema:Person
96 N725538446b6f4bc59defb31b71653a18 rdf:first sg:person.013640554311.55
97 rdf:rest N4d6e50e8d40447249b620c4ef4020d50
98 N72cd36c06add438e909e328b3a247378 schema:familyName Lee
99 schema:givenName Sang-Won
100 rdf:type schema:Person
101 N77250afd670744fda2a6a7584d3fe180 schema:familyName Yu
102 schema:givenName Jeffrey Xu
103 rdf:type schema:Person
104 N7cc54dfd1f62409299a34d4ebb12f7f2 schema:familyName Whang
105 schema:givenName Steven Euijong
106 rdf:type schema:Person
107 N8ce31adc01e44619899e93588d346c72 rdf:first N77250afd670744fda2a6a7584d3fe180
108 rdf:rest Nb514d5f08c9c4bdf9209f1f940fcc0b6
109 N90b0e68f225c45f18696665afa085be7 rdf:first N7cc54dfd1f62409299a34d4ebb12f7f2
110 rdf:rest rdf:nil
111 N9d9cb75ba0e6423e9edd4ce78498707b rdf:first N1ac7bf1399c6422e8c2682e158c4cfd3
112 rdf:rest Nc4d634cb57cb4569a2da32b159b25e52
113 Nb514d5f08c9c4bdf9209f1f940fcc0b6 rdf:first N5ef7fdc03e5c4256ae07b6a5883da885
114 rdf:rest N90b0e68f225c45f18696665afa085be7
115 Nc449e37db74b463992f12c94b12d4c77 schema:name dimensions_id
116 schema:value pub.1131075133
117 rdf:type schema:PropertyValue
118 Nc4d634cb57cb4569a2da32b159b25e52 rdf:first N72cd36c06add438e909e328b3a247378
119 rdf:rest N8ce31adc01e44619899e93588d346c72
120 Ncf3286ff278b4c0ea58aea12171d14c7 rdf:first sg:person.010412660037.85
121 rdf:rest Ned38c8a88cd64d5ea956a2586a0961d7
122 Ndeaffae5bdea4ebe852f58397c1f4df5 schema:isbn 978-3-030-59415-2
123 978-3-030-59416-9
124 schema:name Database Systems for Advanced Applications
125 rdf:type schema:Book
126 Ne6ae7f9bab0a4d4b9c6ccc5cf5b3ef3c schema:name Springer Nature
127 rdf:type schema:Organisation
128 Ned38c8a88cd64d5ea956a2586a0961d7 rdf:first sg:person.012005621037.94
129 rdf:rest N725538446b6f4bc59defb31b71653a18
130 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
131 schema:name Information and Computing Sciences
132 rdf:type schema:DefinedTerm
133 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
134 schema:name Artificial Intelligence and Image Processing
135 rdf:type schema:DefinedTerm
136 sg:person.010412660037.85 schema:affiliation grid-institutes:grid.12527.33
137 schema:familyName Zheng
138 schema:givenName Yanan
139 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010412660037.85
140 rdf:type schema:Person
141 sg:person.012005621037.94 schema:affiliation grid-institutes:grid.411054.5
142 schema:familyName Wang
143 schema:givenName Yan
144 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012005621037.94
145 rdf:type schema:Person
146 sg:person.012303351315.43 schema:affiliation grid-institutes:grid.12527.33
147 schema:familyName Wang
148 schema:givenName Jianmin
149 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012303351315.43
150 rdf:type schema:Person
151 sg:person.013640554311.55 schema:affiliation grid-institutes:grid.12527.33
152 schema:familyName Wen
153 schema:givenName Lijie
154 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013640554311.55
155 rdf:type schema:Person
156 grid-institutes:grid.12527.33 schema:alternateName School of Software, Tsinghua University, Beijing, China
157 schema:name School of Software, Tsinghua University, Beijing, China
158 rdf:type schema:Organization
159 grid-institutes:grid.411054.5 schema:alternateName School of Information, Central University of Finance and Economics, Beijing, China
160 schema:name School of Information, Central University of Finance and Economics, Beijing, China
161 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...