Changes

              
    

          
          
        
        

            
f
1
{
f
1
{
            
2
  "access_rights": "",
2
  "access_rights": "",
            
3
  "accrualPeriodicity": "",
3
  "accrualPeriodicity": "",
            
4
  "author": "Hao Huang",
4
  "author": "Hao Huang",
            
5
  "author_email": "hao.haung@tib.eu",
5
  "author_email": "hao.haung@tib.eu",
            
6
  "citation": [],
6
  "citation": [],
            
7
  "conformsTo": "",
7
  "conformsTo": "",
            
8
  "creator_user_id": "17755db4-395a-4b3b-ac09-e8e3484ca700",
8
  "creator_user_id": "17755db4-395a-4b3b-ac09-e8e3484ca700",
            
9
  "defined_in": "https://doi.org/10.13140/RG.2.2.15682.18889",
9
  "defined_in": "https://doi.org/10.13140/RG.2.2.15682.18889",
            
10
  "doi": "10.57702/9nmma21w",
10
  "doi": "10.57702/9nmma21w",
            
11
  "doi_date_published": null,
11
  "doi_date_published": null,
            
12
  "doi_publisher": "TIB",
12
  "doi_publisher": "TIB",
            
13
  "doi_status": false,
13
  "doi_status": false,
            
14
  "domain": "https://ldm.kisski.de",
14
  "domain": "https://ldm.kisski.de",
            
15
  "end_date": "",
15
  "end_date": "",
            
16
  "extra_authors": [
16
  "extra_authors": [
            
17
    {
17
    {
            
18
      "extra_author": "Emetis Niazmand",
18
      "extra_author": "Emetis Niazmand",
            
19
      "orcid": "0000-0001-8194-8079"
19
      "orcid": "0000-0001-8194-8079"
            
20
    },
20
    },
            
21
    {
21
    {
            
22
      "extra_author": "Maria-Esther Vidal",
22
      "extra_author": "Maria-Esther Vidal",
            
23
      "orcid": "0000-0003-1160-8727"
23
      "orcid": "0000-0003-1160-8727"
            
24
    }
24
    }
            
25
  ],
25
  ],
            
26
  "extras": [
26
  "extras": [
            
27
    {
27
    {
            
28
      "__extras": {
28
      "__extras": {
            
29
        "id": "bd040051-3fb7-4ac7-b93a-fcfee544b0df",
29
        "id": "bd040051-3fb7-4ac7-b93a-fcfee544b0df",
            
30
        "package_id": "6f83604a-9543-4a9c-b49b-450be6a2577c",
30
        "package_id": "6f83604a-9543-4a9c-b49b-450be6a2577c",
            
31
        "state": "active"
31
        "state": "active"
            
32
      },
32
      },
            
33
      "key": "",
33
      "key": "",
            
34
      "value": ""
34
      "value": ""
            
35
    }
35
    }
            
36
  ],
36
  ],
            
37
  "groups": [],
37
  "groups": [],
            
38
  "id": "6f83604a-9543-4a9c-b49b-450be6a2577c",
38
  "id": "6f83604a-9543-4a9c-b49b-450be6a2577c",
            
39
  "isopen": true,
39
  "isopen": true,
            
40
  "landing_page": "",
40
  "landing_page": "",
            
41
  "language": "English",
41
  "language": "English",
            
42
  "license_id": "cc-by",
42
  "license_id": "cc-by",
            
43
  "license_title": "Creative Commons Attribution",
43
  "license_title": "Creative Commons Attribution",
            
44
  "license_url": "http://www.opendefinition.org/licenses/cc-by",
44
  "license_url": "http://www.opendefinition.org/licenses/cc-by",
            
45
  "link_orkg": "http://orkg.org/orkg/resource/R735326",
45
  "link_orkg": "http://orkg.org/orkg/resource/R735326",
            
46
  "maintainer": "Hao Huang",
46
  "maintainer": "Hao Huang",
            
47
  "maintainer_email": "hao.haung@tib.eu",
47
  "maintainer_email": "hao.haung@tib.eu",
            
48
  "metadata_created": "2024-09-16T11:10:13.665218",
48
  "metadata_created": "2024-09-16T11:10:13.665218",
            
n
49
  "metadata_modified": "2024-09-16T11:10:13.902459",
n
49
  "metadata_modified": "2024-09-16T11:11:44.959159",
            
50
  "name": 
50
  "name": 
            
51
nterfactual-prediction-over-knowledge-graphs-for-personal-healthcare",
51
nterfactual-prediction-over-knowledge-graphs-for-personal-healthcare",
            
52
  "notes": "https://youtu.be/vkF3fpCZLhk Artificial intelligence (AI) 
52
  "notes": "https://youtu.be/vkF3fpCZLhk Artificial intelligence (AI) 
            
53
has become an invaluable tool in health-care for disease prediction 
53
has become an invaluable tool in health-care for disease prediction 
            
54
and diagnosis. Despite their predictive accuracy, AI models may ignore 
54
and diagnosis. Despite their predictive accuracy, AI models may ignore 
            
55
causal relationships between patient characteristics (demographic or 
55
causal relationships between patient characteristics (demographic or 
            
56
clinical). As a result, although AI models capture associative 
56
clinical). As a result, although AI models capture associative 
            
57
patterns, ignoring the causal relationships of predicted 
57
patterns, ignoring the causal relationships of predicted 
            
58
characteristics limits their ability to perform counterfactual 
58
characteristics limits their ability to perform counterfactual 
            
59
reasoning about the predicted characteristics of a patient when 
59
reasoning about the predicted characteristics of a patient when 
            
60
dependent characteristics change. This limitation of AI models can 
60
dependent characteristics change. This limitation of AI models can 
            
61
affect the understanding of predicted outcomes. We have proposed 
61
affect the understanding of predicted outcomes. We have proposed 
            
62
hybrid AI methods that combine symbolic reasoning over knowledge 
62
hybrid AI methods that combine symbolic reasoning over knowledge 
            
63
graphs (KGs), large language models (LLMs), and causal reasoning 
63
graphs (KGs), large language models (LLMs), and causal reasoning 
            
64
techniques to infer causal relationships between patients' properties. 
64
techniques to infer causal relationships between patients' properties. 
            
65
As a result, a causal model is learned, enabling counterfactual 
65
As a result, a causal model is learned, enabling counterfactual 
            
66
prediction to support clinical decisions. We apply these AI methods to 
66
prediction to support clinical decisions. We apply these AI methods to 
            
67
predict the counterfactuals of biomarker results in non-small cell 
67
predict the counterfactuals of biomarker results in non-small cell 
            
68
lung cancer (NSCLC) patients under hypothetical treatments with 
68
lung cancer (NSCLC) patients under hypothetical treatments with 
            
69
different smoking habits. We have created synthetic datasets based on 
69
different smoking habits. We have created synthetic datasets based on 
            
70
clinical records of NSCLC patients to evaluate the performance of the 
70
clinical records of NSCLC patients to evaluate the performance of the 
            
71
proposed methods. The observed results suggest that our methods are 
71
proposed methods. The observed results suggest that our methods are 
            
72
competitive with baseline methods in causal relationship discovery and 
72
competitive with baseline methods in causal relationship discovery and 
            
73
counterfactual prediction. CCS CONCEPTS \u2022 Applied computing 
73
counterfactual prediction. CCS CONCEPTS \u2022 Applied computing 
            
74
\u2192 Health care information systems; \u2022 Computing methodologies 
74
\u2192 Health care information systems; \u2022 Computing methodologies 
            
75
\u2192 Artificial intelligence.",
75
\u2192 Artificial intelligence.",
            
n
76
  "num_resources": 0,
n
76
  "num_resources": 1,
            
77
  "num_tags": 3,
77
  "num_tags": 3,
            
78
  "orcid": "0000-0003-3117-0881",
78
  "orcid": "0000-0003-3117-0881",
            
79
  "organization": {
79
  "organization": {
            
80
    "approval_status": "approved",
80
    "approval_status": "approved",
            
81
    "created": "2017-11-23T17:30:37.757128",
81
    "created": "2017-11-23T17:30:37.757128",
            
82
    "description": "The German National Library of Science and 
82
    "description": "The German National Library of Science and 
            
83
Technology, abbreviated TIB, is the national library of the Federal 
83
Technology, abbreviated TIB, is the national library of the Federal 
            
84
Republic of Germany for all fields of engineering, technology, and the 
84
Republic of Germany for all fields of engineering, technology, and the 
            
85
natural sciences.",
85
natural sciences.",
            
86
    "id": "0c5362f5-b99e-41db-8256-3d0d7549bf4d",
86
    "id": "0c5362f5-b99e-41db-8256-3d0d7549bf4d",
            
87
    "image_url": 
87
    "image_url": 
            
88
3conf/ext/tib_tmpl_bootstrap/Resources/Public/images/TIB_Logo_en.png",
88
3conf/ext/tib_tmpl_bootstrap/Resources/Public/images/TIB_Logo_en.png",
            
89
    "is_organization": true,
89
    "is_organization": true,
            
90
    "name": "tib",
90
    "name": "tib",
            
91
    "state": "active",
91
    "state": "active",
            
92
    "title": "TIB",
92
    "title": "TIB",
            
93
    "type": "organization"
93
    "type": "organization"
            
94
  },
94
  },
            
95
  "owner_org": "0c5362f5-b99e-41db-8256-3d0d7549bf4d",
95
  "owner_org": "0c5362f5-b99e-41db-8256-3d0d7549bf4d",
            
96
  "page": "",
96
  "page": "",
            
97
  "private": false,
97
  "private": false,
            
98
  "relationships_as_object": [],
98
  "relationships_as_object": [],
            
99
  "relationships_as_subject": [],
99
  "relationships_as_subject": [],
            
t
100
  "resources": [],
t
100
  "resources": [
            
101
    {
            
102
      "auto_update_last_update": "",
            
103
      "cache_last_updated": null,
            
104
      "cache_url": null,
            
105
      "created": "2024-09-16T11:11:44.970346",
            
106
      "description": "",
            
107
      "format": "ZIP",
            
108
      "hash": "",
            
109
      "id": "2b9aff9c-16b4-4591-9ab9-577f3d27ff50",
            
110
      "language": "",
            
111
      "last_modified": "2024-09-16T11:11:44.947735",
            
112
      "media": "",
            
113
      "metadata_modified": "2024-09-16T11:11:44.962458",
            
114
      "mimetype": "application/zip",
            
115
      "mimetype_inner": null,
            
116
      "name": "Workshop Data Sources",
            
117
      "package_id": "6f83604a-9543-4a9c-b49b-450be6a2577c",
            
118
      "position": 0,
            
119
      "resource_type": null,
            
120
      "rights": "",
            
121
      "size": 789346,
            
122
      "state": "active",
            
123
      "url": 
            
124
e/2b9aff9c-16b4-4591-9ab9-577f3d27ff50/download/kdd-healthcareai.zip",
            
125
      "url_type": "upload"
            
126
    }
            
127
  ],
            
101
  "services_used_list": "",
128
  "services_used_list": "",
            
102
  "spatial": "",
129
  "spatial": "",
            
103
  "spatial_resolution": "",
130
  "spatial_resolution": "",
            
104
  "start_date": "",
131
  "start_date": "",
            
105
  "state": "draft",
132
  "state": "draft",
            
106
  "tags": [
133
  "tags": [
            
107
    {
134
    {
            
108
      "display_name": "Artificial intelligence",
135
      "display_name": "Artificial intelligence",
            
109
      "id": "19c5fcb5-fa24-442c-a559-a9fff2d58735",
136
      "id": "19c5fcb5-fa24-442c-a559-a9fff2d58735",
            
110
      "name": "Artificial intelligence",
137
      "name": "Artificial intelligence",
            
111
      "state": "active",
138
      "state": "active",
            
112
      "vocabulary_id": null
139
      "vocabulary_id": null
            
113
    },
140
    },
            
114
    {
141
    {
            
115
      "display_name": "Knowledge Graph",
142
      "display_name": "Knowledge Graph",
            
116
      "id": "12b0c897-01cf-4275-99e8-21d85c09a38d",
143
      "id": "12b0c897-01cf-4275-99e8-21d85c09a38d",
            
117
      "name": "Knowledge Graph",
144
      "name": "Knowledge Graph",
            
118
      "state": "active",
145
      "state": "active",
            
119
      "vocabulary_id": null
146
      "vocabulary_id": null
            
120
    },
147
    },
            
121
    {
148
    {
            
122
      "display_name": "healthcare",
149
      "display_name": "healthcare",
            
123
      "id": "e187bf66-1085-49ef-a3dc-34dc68276244",
150
      "id": "e187bf66-1085-49ef-a3dc-34dc68276244",
            
124
      "name": "healthcare",
151
      "name": "healthcare",
            
125
      "state": "active",
152
      "state": "active",
            
126
      "vocabulary_id": null
153
      "vocabulary_id": null
            
127
    }
154
    }
            
128
  ],
155
  ],
            
129
  "temporal_resolution": "",
156
  "temporal_resolution": "",
            
130
  "title": "Hybrid AI Approach for Counterfactual Prediction over 
157
  "title": "Hybrid AI Approach for Counterfactual Prediction over 
            
131
Knowledge Graphs for Personal Healthcare",
158
Knowledge Graphs for Personal Healthcare",
            
132
  "type": "dataset",
159
  "type": "dataset",
            
133
  "url": 
160
  "url": 
            
134
nterfactual_Prediction_over_Knowledge_Graphs_for_Personal_Healthcare",
161
nterfactual_Prediction_over_Knowledge_Graphs_for_Personal_Healthcare",
            
135
  "version": "",
162
  "version": "",
            
136
  "version_note": ""
163
  "version_note": ""
            
137
}
164
}