Explaining the Unexplained: Revealing Hidden Correlations for Better Interpretability Article Swipe

PDF

Wendong Jiang , Chih‐Yung Chang , Show-Jane Yen , Diptendu Sinha Roy ·

YOU? · · 2024 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2412.01365

Deep learning has achieved remarkable success in processing and managing unstructured data. However, its "black box" nature imposes significant limitations, particularly in sensitive application domains. While existing interpretable machine learning methods address some of these issues, they often fail to adequately consider feature correlations and provide insufficient evaluation of model decision paths. To overcome these challenges, this paper introduces Real Explainer (RealExp), an interpretability computation method that decouples the Shapley Value into individual feature importance and feature correlation importance. By incorporating feature similarity computations, RealExp enhances interpretability by precisely quantifying both individual feature contributions and their interactions, leading to more reliable and nuanced explanations. Additionally, this paper proposes a novel interpretability evaluation criterion focused on elucidating the decision paths of deep learning models, going beyond traditional accuracy-based metrics. Experimental validations on two unstructured data tasks -- image classification and text sentiment analysis -- demonstrate that RealExp significantly outperforms existing methods in interpretability. Case studies further illustrate its practical value: in image classification, RealExp aids in selecting suitable pre-trained models for specific tasks from an interpretability perspective; in text classification, it enables the optimization of models and approximates the performance of a fine-tuned GPT-Ada model using traditional bag-of-words approaches.

Related Topics

Computer Science

Artificial Intelligence

Economics

Concepts

Interpretability Econometrics Computer science Artificial intelligence Economics

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2412.01365
PDF: https://arxiv.org/pdf/2412.01365
OA Status: green
Related Works: 10
OpenAlex ID: https://openalex.org/W4405034271

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4405034271

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2412.01365

Digital Object Identifier
Title: Explaining the Unexplained: Revealing Hidden Correlations for Better Interpretability

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2024

Year of publication
Publication date: 2024-12-02

Full publication date if available
Authors: Wendong Jiang, Chih‐Yung Chang, Show-Jane Yen, Diptendu Sinha Roy

List of authors in order
Landing page: https://arxiv.org/abs/2412.01365

Publisher landing page
PDF URL: https://arxiv.org/pdf/2412.01365

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2412.01365

Direct OA link when available
Concepts: Interpretability, Econometrics, Computer science, Artificial intelligence, Economics

Top concepts (fields/topics) attached by OpenAlex
Cited by: 0

Total citation count in OpenAlex
Related works (count): 10

Other works algorithmically related by OpenAlex

Full payload

id	https://openalex.org/W4405034271
doi	https://doi.org/10.48550/arxiv.2412.01365
ids.doi	https://doi.org/10.48550/arxiv.2412.01365
ids.openalex	https://openalex.org/W4405034271
fwci
type	preprint
title	Explaining the Unexplained: Revealing Hidden Correlations for Better Interpretability
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
topics[0].id	https://openalex.org/T10028
topics[0].field.id	https://openalex.org/fields/17
topics[0].field.display_name	Computer Science
topics[0].score	0.8079000115394592
topics[0].domain.id	https://openalex.org/domains/3
topics[0].domain.display_name	Physical Sciences
topics[0].subfield.id	https://openalex.org/subfields/1702
topics[0].subfield.display_name	Artificial Intelligence
topics[0].display_name	Topic Modeling
is_xpac	False
apc_list
apc_paid
concepts[0].id	https://openalex.org/C2781067378
concepts[0].level	2
concepts[0].score	0.8964316248893738
concepts[0].wikidata	https://www.wikidata.org/wiki/Q17027399
concepts[0].display_name	Interpretability
concepts[1].id	https://openalex.org/C149782125
concepts[1].level	1
concepts[1].score	0.3827180564403534
concepts[1].wikidata	https://www.wikidata.org/wiki/Q160039
concepts[1].display_name	Econometrics
concepts[2].id	https://openalex.org/C41008148
concepts[2].level	0
concepts[2].score	0.33376359939575195
concepts[2].wikidata	https://www.wikidata.org/wiki/Q21198
concepts[2].display_name	Computer science
concepts[3].id	https://openalex.org/C154945302
concepts[3].level	1
concepts[3].score	0.307199090719223
concepts[3].wikidata	https://www.wikidata.org/wiki/Q11660
concepts[3].display_name	Artificial intelligence
concepts[4].id	https://openalex.org/C162324750
concepts[4].level	0
concepts[4].score	0.2785315215587616
concepts[4].wikidata	https://www.wikidata.org/wiki/Q8134
concepts[4].display_name	Economics
keywords[0].id	https://openalex.org/keywords/interpretability
keywords[0].score	0.8964316248893738
keywords[0].display_name	Interpretability
keywords[1].id	https://openalex.org/keywords/econometrics
keywords[1].score	0.3827180564403534
keywords[1].display_name	Econometrics
keywords[2].id	https://openalex.org/keywords/computer-science
keywords[2].score	0.33376359939575195
keywords[2].display_name	Computer science
keywords[3].id	https://openalex.org/keywords/artificial-intelligence
keywords[3].score	0.307199090719223
keywords[3].display_name	Artificial intelligence
keywords[4].id	https://openalex.org/keywords/economics
keywords[4].score	0.2785315215587616
keywords[4].display_name	Economics
language	en
locations[0].id	pmh:oai:arXiv.org:2412.01365
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2412.01365
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2412.01365
locations[1].id	doi:10.48550/arxiv.2412.01365
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2412.01365
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5047519765
authorships[0].author.orcid
authorships[0].author.display_name	Wendong Jiang
authorships[0].author_position	first
authorships[0].raw_author_name	Jiang, Wen-Dong
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5002582301
authorships[1].author.orcid	https://orcid.org/0000-0002-0672-5593
authorships[1].author.display_name	Chih‐Yung Chang
authorships[1].author_position	middle
authorships[1].raw_author_name	Chang, Chih-Yung
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5037108143
authorships[2].author.orcid	https://orcid.org/0009-0006-5100-5030
authorships[2].author.display_name	Show-Jane Yen
authorships[2].author_position	middle
authorships[2].raw_author_name	Yen, Show-Jane
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5011943818
authorships[3].author.orcid	https://orcid.org/0000-0001-9731-2534
authorships[3].author.display_name	Diptendu Sinha Roy
authorships[3].author_position	last
authorships[3].raw_author_name	Roy, Diptendu Sinha
authorships[3].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2412.01365
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-10T00:00:00
display_name	Explaining the Unexplained: Revealing Hidden Correlations for Better Interpretability
has_fulltext	False
is_retracted	False
updated_date	2025-11-06T06:51:31.235846
primary_topic.id	https://openalex.org/T10028
primary_topic.field.id	https://openalex.org/fields/17
primary_topic.field.display_name	Computer Science
primary_topic.score	0.8079000115394592
primary_topic.domain.id	https://openalex.org/domains/3
primary_topic.domain.display_name	Physical Sciences
primary_topic.subfield.id	https://openalex.org/subfields/1702
primary_topic.subfield.display_name	Artificial Intelligence
primary_topic.display_name	Topic Modeling
related_works	https://openalex.org/W4391375266, https://openalex.org/W2899084033, https://openalex.org/W2748952813, https://openalex.org/W2905433371, https://openalex.org/W2888392564, https://openalex.org/W4310278675, https://openalex.org/W4388422664, https://openalex.org/W4390569940, https://openalex.org/W4361193272, https://openalex.org/W2963326959
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2412.01365
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2412.01365
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2412.01365
primary_location.id	pmh:oai:arXiv.org:2412.01365
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2412.01365
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2412.01365
publication_date	2024-12-02
publication_year	2024
referenced_works_count	0
abstract_inverted_index.a	108, 190
abstract_inverted_index.--	135, 142
abstract_inverted_index.By	79
abstract_inverted_index.To	52
abstract_inverted_index.an	62, 173
abstract_inverted_index.by	87
abstract_inverted_index.in	6, 21, 150, 159, 164, 176
abstract_inverted_index.it	179
abstract_inverted_index.of	33, 48, 119, 183, 189
abstract_inverted_index.on	114, 130
abstract_inverted_index.to	39, 98
abstract_inverted_index.and	8, 44, 75, 94, 101, 138, 185
abstract_inverted_index.for	169
abstract_inverted_index.has	2
abstract_inverted_index.its	13, 156
abstract_inverted_index.the	68, 116, 181, 187
abstract_inverted_index.two	131
abstract_inverted_index.Case	152
abstract_inverted_index.Deep	0
abstract_inverted_index.Real	59
abstract_inverted_index.aids	163
abstract_inverted_index.both	90
abstract_inverted_index.box"	15
abstract_inverted_index.data	133
abstract_inverted_index.deep	120
abstract_inverted_index.fail	38
abstract_inverted_index.from	172
abstract_inverted_index.into	71
abstract_inverted_index.more	99
abstract_inverted_index.some	32
abstract_inverted_index.text	139, 177
abstract_inverted_index.that	66, 144
abstract_inverted_index.they	36
abstract_inverted_index.this	56, 105
abstract_inverted_index.Value	70
abstract_inverted_index.While	25
abstract_inverted_index.data.	11
abstract_inverted_index.going	123
abstract_inverted_index.image	136, 160
abstract_inverted_index.model	49, 193
abstract_inverted_index.novel	109
abstract_inverted_index.often	37
abstract_inverted_index.paper	57, 106
abstract_inverted_index.paths	118
abstract_inverted_index.tasks	134, 171
abstract_inverted_index.their	95
abstract_inverted_index.these	34, 54
abstract_inverted_index.using	194
abstract_inverted_index."black	14
abstract_inverted_index.beyond	124
abstract_inverted_index.method	65
abstract_inverted_index.models	168, 184
abstract_inverted_index.nature	16
abstract_inverted_index.paths.	51
abstract_inverted_index.value:	158
abstract_inverted_index.GPT-Ada	192
abstract_inverted_index.RealExp	84, 145, 162
abstract_inverted_index.Shapley	69
abstract_inverted_index.address	31
abstract_inverted_index.enables	180
abstract_inverted_index.feature	42, 73, 76, 81, 92
abstract_inverted_index.focused	113
abstract_inverted_index.further	154
abstract_inverted_index.imposes	17
abstract_inverted_index.issues,	35
abstract_inverted_index.leading	97
abstract_inverted_index.machine	28
abstract_inverted_index.methods	30, 149
abstract_inverted_index.models,	122
abstract_inverted_index.nuanced	102
abstract_inverted_index.provide	45
abstract_inverted_index.studies	153
abstract_inverted_index.success	5
abstract_inverted_index.However,	12
abstract_inverted_index.achieved	3
abstract_inverted_index.analysis	141
abstract_inverted_index.consider	41
abstract_inverted_index.decision	50, 117
abstract_inverted_index.domains.	24
abstract_inverted_index.enhances	85
abstract_inverted_index.existing	26, 148
abstract_inverted_index.learning	1, 29, 121
abstract_inverted_index.managing	9
abstract_inverted_index.metrics.	127
abstract_inverted_index.overcome	53
abstract_inverted_index.proposes	107
abstract_inverted_index.reliable	100
abstract_inverted_index.specific	170
abstract_inverted_index.suitable	166
abstract_inverted_index.Explainer	60
abstract_inverted_index.criterion	112
abstract_inverted_index.decouples	67
abstract_inverted_index.practical	157
abstract_inverted_index.precisely	88
abstract_inverted_index.selecting	165
abstract_inverted_index.sensitive	22
abstract_inverted_index.sentiment	140
abstract_inverted_index.(RealExp),	61
abstract_inverted_index.adequately	40
abstract_inverted_index.evaluation	47, 111
abstract_inverted_index.fine-tuned	191
abstract_inverted_index.illustrate	155
abstract_inverted_index.importance	74
abstract_inverted_index.individual	72, 91
abstract_inverted_index.introduces	58
abstract_inverted_index.processing	7
abstract_inverted_index.remarkable	4
abstract_inverted_index.similarity	82
abstract_inverted_index.application	23
abstract_inverted_index.approaches.	197
abstract_inverted_index.challenges,	55
abstract_inverted_index.computation	64
abstract_inverted_index.correlation	77
abstract_inverted_index.demonstrate	143
abstract_inverted_index.elucidating	115
abstract_inverted_index.importance.	78
abstract_inverted_index.outperforms	147
abstract_inverted_index.performance	188
abstract_inverted_index.pre-trained	167
abstract_inverted_index.quantifying	89
abstract_inverted_index.significant	18
abstract_inverted_index.traditional	125, 195
abstract_inverted_index.validations	129
abstract_inverted_index.Experimental	128
abstract_inverted_index.approximates	186
abstract_inverted_index.bag-of-words	196
abstract_inverted_index.correlations	43
abstract_inverted_index.insufficient	46
abstract_inverted_index.limitations,	19
abstract_inverted_index.optimization	182
abstract_inverted_index.particularly	20
abstract_inverted_index.perspective;	175
abstract_inverted_index.unstructured	10, 132
abstract_inverted_index.Additionally,	104
abstract_inverted_index.computations,	83
abstract_inverted_index.contributions	93
abstract_inverted_index.explanations.	103
abstract_inverted_index.incorporating	80
abstract_inverted_index.interactions,	96
abstract_inverted_index.interpretable	27
abstract_inverted_index.significantly	146
abstract_inverted_index.accuracy-based	126
abstract_inverted_index.classification	137
abstract_inverted_index.classification,	161, 178
abstract_inverted_index.interpretability	63, 86, 110, 174
abstract_inverted_index.interpretability.	151
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	4
citation_normalized_percentile