RKB109 commited on
Commit
5aed858
·
verified ·
1 Parent(s): 5ae0b29

Publish artifacts for knowledge-graph-risk-engine-20260808

Browse files
Files changed (5) hide show
  1. README.md +59 -0
  2. data/test.jsonl +4 -0
  3. data/train.jsonl +14 -0
  4. project.json +52 -0
  5. sources.json +12 -0
README.md ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ language:
4
+ - en
5
+ pretty_name: Knowledge Graph Risk Engine Synthetic Evaluation Set
6
+ size_categories:
7
+ - n<1K
8
+ task_categories:
9
+ - feature-extraction
10
+ tags:
11
+ - synthetic
12
+ - knowledge-graphs
13
+ - evaluation
14
+ - token-classification
15
+ - feature-extraction
16
+ - question-answering
17
+ - sentence-similarity
18
+ configs:
19
+ - config_name: default
20
+ data_files:
21
+ - split: train
22
+ path: data/train.jsonl
23
+ - split: test
24
+ path: data/test.jsonl
25
+ ---
26
+
27
+ # Knowledge Graph Risk Engine Synthetic Dataset
28
+
29
+ ## Summary
30
+
31
+ This dataset contains 14 training examples and 4
32
+ held-out examples for **Risk teams need relationship-level explanations instead of opaque entity scores.**
33
+
34
+ Every record is synthetic and includes:
35
+
36
+ - `input`: query, event, or feature description
37
+ - `label`: expected class, route, relation, or evidence category
38
+ - `context`: synthetic supporting context
39
+ - `source`: fictional source identifier
40
+ - `variant`: generation pattern
41
+ - `synthetic`: always `true`
42
+
43
+ ## Uses
44
+
45
+ - Reproducible unit and integration tests
46
+ - Baseline model training
47
+ - Evaluation harness development
48
+ - Schema and architecture demonstrations
49
+
50
+ ## Limitations
51
+
52
+ All entities are fictional. Real identity or financial data requires governance, consent, and bias review.
53
+
54
+ This dataset does not represent real users, patients, customers, production
55
+ traffic, or licensed media. It must not be presented as real-world evidence.
56
+
57
+ ## Related Model
58
+
59
+ [RKB109/knowledge-graph-risk-engine-20260808-model](https://huggingface.co/RKB109/knowledge-graph-risk-engine-20260808-model)
data/test.jsonl ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {"id":"knowledge-graph-risk-2-1","input":"Which account paid vendor beta?","label":"transaction","context":"account-17 paid vendor-beta","source":"graph-02","variant":"direct","synthetic":true}
2
+ {"id":"knowledge-graph-risk-3-2","input":"In an operations review, Where is warehouse north located?","label":"location","context":"warehouse-north located-in region-east","source":"graph-03","variant":"operations","synthetic":true}
3
+ {"id":"knowledge-graph-risk-4-3","input":"For an evaluation case, Which company owns vendor gamma?","label":"ownership","context":"company-delta owns vendor-gamma","source":"graph-04","variant":"evaluation","synthetic":true}
4
+ {"id":"knowledge-graph-risk-6-1","input":"What region contains facility blue?","label":"location","context":"facility-blue located-in region-west","source":"graph-06","variant":"direct","synthetic":true}
data/train.jsonl ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"id":"knowledge-graph-risk-1-1","input":"Who controls supplier alpha?","label":"ownership","context":"company-alpha controls supplier-alpha","source":"graph-01","variant":"direct","synthetic":true}
2
+ {"id":"knowledge-graph-risk-1-2","input":"In an operations review, Who controls supplier alpha?","label":"ownership","context":"company-alpha controls supplier-alpha","source":"graph-01","variant":"operations","synthetic":true}
3
+ {"id":"knowledge-graph-risk-1-3","input":"For an evaluation case, Who controls supplier alpha?","label":"ownership","context":"company-alpha controls supplier-alpha","source":"graph-01","variant":"evaluation","synthetic":true}
4
+ {"id":"knowledge-graph-risk-2-2","input":"In an operations review, Which account paid vendor beta?","label":"transaction","context":"account-17 paid vendor-beta","source":"graph-02","variant":"operations","synthetic":true}
5
+ {"id":"knowledge-graph-risk-2-3","input":"For an evaluation case, Which account paid vendor beta?","label":"transaction","context":"account-17 paid vendor-beta","source":"graph-02","variant":"evaluation","synthetic":true}
6
+ {"id":"knowledge-graph-risk-3-1","input":"Where is warehouse north located?","label":"location","context":"warehouse-north located-in region-east","source":"graph-03","variant":"direct","synthetic":true}
7
+ {"id":"knowledge-graph-risk-3-3","input":"For an evaluation case, Where is warehouse north located?","label":"location","context":"warehouse-north located-in region-east","source":"graph-03","variant":"evaluation","synthetic":true}
8
+ {"id":"knowledge-graph-risk-4-1","input":"Which company owns vendor gamma?","label":"ownership","context":"company-delta owns vendor-gamma","source":"graph-04","variant":"direct","synthetic":true}
9
+ {"id":"knowledge-graph-risk-4-2","input":"In an operations review, Which company owns vendor gamma?","label":"ownership","context":"company-delta owns vendor-gamma","source":"graph-04","variant":"operations","synthetic":true}
10
+ {"id":"knowledge-graph-risk-5-1","input":"What transfer connects account 17 and supplier alpha?","label":"transaction","context":"account-17 transferred-to supplier-alpha","source":"graph-05","variant":"direct","synthetic":true}
11
+ {"id":"knowledge-graph-risk-5-2","input":"In an operations review, What transfer connects account 17 and supplier alpha?","label":"transaction","context":"account-17 transferred-to supplier-alpha","source":"graph-05","variant":"operations","synthetic":true}
12
+ {"id":"knowledge-graph-risk-5-3","input":"For an evaluation case, What transfer connects account 17 and supplier alpha?","label":"transaction","context":"account-17 transferred-to supplier-alpha","source":"graph-05","variant":"evaluation","synthetic":true}
13
+ {"id":"knowledge-graph-risk-6-2","input":"In an operations review, What region contains facility blue?","label":"location","context":"facility-blue located-in region-west","source":"graph-06","variant":"operations","synthetic":true}
14
+ {"id":"knowledge-graph-risk-6-3","input":"For an evaluation case, What region contains facility blue?","label":"location","context":"facility-blue located-in region-west","source":"graph-06","variant":"evaluation","synthetic":true}
project.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "Knowledge Graph Risk Engine",
3
+ "problem": "Risk teams need relationship-level explanations instead of opaque entity scores.",
4
+ "domain": "knowledge-graphs",
5
+ "architecture": "graph",
6
+ "hugging_face_tasks": [
7
+ "token-classification",
8
+ "feature-extraction",
9
+ "question-answering",
10
+ "sentence-similarity"
11
+ ],
12
+ "recommended_stack": [
13
+ "FastAPI for entity and evidence APIs",
14
+ "Neo4j Community or PostgreSQL recursive queries",
15
+ "Sentence Transformers for entity resolution",
16
+ "NetworkX for local graph validation",
17
+ "Kafka-compatible event ingestion",
18
+ "OpenTelemetry for lineage and query traces"
19
+ ],
20
+ "real_world_data_sources": [
21
+ {
22
+ "name": "SEC EDGAR submissions API",
23
+ "url": "https://data.sec.gov/submissions/CIK0000320193.json",
24
+ "purpose": "Public company and filing relationships"
25
+ },
26
+ {
27
+ "name": "GLEIF LEI API",
28
+ "url": "https://api.gleif.org/api/v1/lei-records?page[size]=5",
29
+ "purpose": "Public legal-entity identifiers and relationships"
30
+ }
31
+ ],
32
+ "job_description_skills": [
33
+ "Entity resolution and relation extraction",
34
+ "Knowledge-graph modeling and path queries",
35
+ "Graph-based explainability and provenance",
36
+ "Streaming ingestion and schema evolution",
37
+ "Risk-model evaluation and data quality controls"
38
+ ],
39
+ "impact_targets": [
40
+ "Reach entity-resolution precision >= 0.95 on reviewed pairs",
41
+ "Return evidence paths for 100% of emitted risk flags",
42
+ "Process 10,000 relationship events per minute in load tests",
43
+ "Detect schema and orphan-node regressions in CI"
44
+ ],
45
+ "baseline_evaluation": {
46
+ "test_examples": 4,
47
+ "accuracy": 1,
48
+ "synthetic_evaluation": true
49
+ },
50
+ "estimated_delivery": "8-12 weeks for one engineer",
51
+ "generated_baseline_is_production_ready": false
52
+ }
sources.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "name": "SEC EDGAR submissions API",
4
+ "url": "https://data.sec.gov/submissions/CIK0000320193.json",
5
+ "purpose": "Public company and filing relationships"
6
+ },
7
+ {
8
+ "name": "GLEIF LEI API",
9
+ "url": "https://api.gleif.org/api/v1/lei-records?page[size]=5",
10
+ "purpose": "Public legal-entity identifiers and relationships"
11
+ }
12
+ ]