Kyeongmin/C.E.L_Slide_test2
1
0
Fork 0
Code Issues 6 Pull Requests Actions Packages Projects Releases Wiki Activity

wip: phase_z2 evidence 파이프라인 + matching 실험(phase2~26) + 프론트 trace 패널 진행분 스냅샷

Browse Source 
- src: phase_z2 composition/mapper/pipeline/placement_planner/retry, ai_fallback(prompts/schema/validate), mdx_text_atoms 신규
- Front: PipelineTracePanel 신규, FramePanel/SlideCanvas/Home/designAgentApi 등 갱신 + 테스트 4종 추가
- templates/phase_z2: catalog(component_expansion_registry, node_slot_mapping 신규), frames, families, slide_base 갱신
- tests/matching: phase2~26 매칭 실험 스크립트·리포트·온톨로지 전체 (미커밋 진행분)
- tests: b4_v4 evidence, task5~28.5 시리즈, regression(imp95 baseline) 등 신규 테스트 대량 추가
- docs/reference: MDX 구조 인벤토리, MDX→Frame 구조 계약 문서
- scripts: mdx 계약/parity/coverage/viewport 체크, gitea comment, run sync 유틸
- .gitignore: tmp*.json, chromedriver, .orchestrator, *.pkl, Front_test* 등 임시/스냅샷 제외

미완성 작업의 보존용 스냅샷 커밋 (2026-07-02)

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Kyeongmin
2026-07-02 17:03:42 +09:00
co-authored by Claude Opus 4.8
parent 97b7833a1b
commit b836e79ee1
527 changed files with 673036 additions and 717 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/matching/v2_semantic_rerank_result.yaml
+618
View File
@@ -0,0 +1,618 @@
meta:
pipeline_step: 8.v2
description: V1 Top-K 후보를 ko-sroberta cosine 으로 재정렬 (캐스케이드 rerank 1단)
model: jhgan/ko-sroberta-multitask
similarity: cosine
top_k: 5
mdx_summary_spec: title + 첫 일반 문단(≤120자) + slot label Top-5 join (detect_mdx.build_summary)
frame_source: analysis.md "내용" 섹션 (phase_common.load_32_frames)
v1_source: mdx_matching_result.yaml rank_by_matching_score
holdout_sections:
- 01-1
- 02-1
- 02-2.1
- 04-1
- 04-2.1
- 04-2.2
answer_map:
01-2: 18
02-2.2: 14
03-1: 13
03-2: 29
lock_snapshot:
timestamp: '2026-04-29T11:19:08'
model: jhgan/ko-sroberta-multitask
top_k: 5
files:
pipeline_08_v2_semantic_rerank.py: d38474c192eeca0994d24f8d9de41c0dd34291683caae1028100917f9ca313bd
embeddings.py: acc50f3f93935061fb146c3b15e30bc9734b4c96c53f82a603bc758944b50c9c
detect_mdx.py: 71192216908246f101ddf5d49fe044266bddd5019ea60a8ac3fbd9c1ea215fbc
phase_common.py: b778840c8e246d0e5210359263e4a59ef6acfa6f3c9a8a1cc110f18506cc4f94
pipeline_01_extract_nodes.py: 26bc7fb52c2ba34459ea7294f191e5aff900618c62d10467535f5ab2eade09dd
synonyms.yaml: 57bbc60d0ca8a95c445e3e0acf1af4e1cff101957ae8c1759311ec827c961893
mdx_matching_result.yaml: f50385fe1eed619a6206b316e1dd4810b0ab489fd8f098c07279d53234f1e1fb
principle:
- V1 baseline 고정 — 재정렬 대상만 Top-K 로 한정
- 가중합 금지 — 순수 semantic rerank
- Holdout 성적을 설계 근거로 사용하지 않음
mdx_sections:
01-2:
section_type: target
answer_frame_number: 18
mdx_title: 2. 용어간 상호관계
mdx_summary: 2. 용어간 상호관계. ![DX와 핵심기술간 상호관계](/assets/images/DX1.png). DX / BIM
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281195'
frame_number: 18
v1_score: 0.929
- rank: 2
frame_id: '1171281182'
frame_number: 11
v1_score: 0.32
- rank: 3
frame_id: '1171281208'
frame_number: 27
v1_score: 0.31
- rank: 4
frame_id: '1171281172'
frame_number: 1
v1_score: 0.3
- rank: 5
frame_id: '1171281180'
frame_number: 9
v1_score: 0.294
v2_rerank:
- frame_id: '1171281195'
frame_number: 18
v1_rank: 1
v1_score: 0.929
semantic_score: 0.7295
v2_rank: 1
- frame_id: '1171281208'
frame_number: 27
v1_rank: 3
v1_score: 0.31
semantic_score: 0.6442
v2_rank: 2
- frame_id: '1171281182'
frame_number: 11
v1_rank: 2
v1_score: 0.32
semantic_score: 0.5853
v2_rank: 3
- frame_id: '1171281172'
frame_number: 1
v1_rank: 4
v1_score: 0.3
semantic_score: 0.5741
v2_rank: 4
- frame_id: '1171281180'
frame_number: 9
v1_rank: 5
v1_score: 0.294
semantic_score: 0.5556
v2_rank: 5
02-2.2:
section_type: target
answer_frame_number: 14
mdx_title: 2.2 DX 시행 주체별 기대효과
mdx_summary: 2.2 DX 시행 주체별 기대효과. 발주자 / 시공자 / 설계자
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281191'
frame_number: 14
v1_score: 0.829
- rank: 2
frame_id: '1171281180'
frame_number: 9
v1_score: 0.445
- rank: 3
frame_id: '1171281176'
frame_number: 5
v1_score: 0.413
- rank: 4
frame_id: '1171281189'
frame_number: 12
v1_score: 0.379
- rank: 5
frame_id: '1171281174'
frame_number: 3
v1_score: 0.344
v2_rerank:
- frame_id: '1171281191'
frame_number: 14
v1_rank: 1
v1_score: 0.829
semantic_score: 0.695
v2_rank: 1
- frame_id: '1171281189'
frame_number: 12
v1_rank: 4
v1_score: 0.379
semantic_score: 0.5615
v2_rank: 2
- frame_id: '1171281174'
frame_number: 3
v1_rank: 5
v1_score: 0.344
semantic_score: 0.5266
v2_rank: 3
- frame_id: '1171281180'
frame_number: 9
v1_rank: 2
v1_score: 0.445
semantic_score: 0.5043
v2_rank: 4
- frame_id: '1171281176'
frame_number: 5
v1_rank: 3
v1_score: 0.413
semantic_score: 0.4486
v2_rank: 5
03-1:
section_type: target
answer_frame_number: 13
mdx_title: 1. DX 시행을 위한 필수 요건
mdx_summary: 1. DX 시행을 위한 필수 요건. 기술 / 사람 / 자연
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281190'
frame_number: 13
v1_score: 0.563
- rank: 2
frame_id: '1171281192'
frame_number: 15
v1_score: 0.517
- rank: 3
frame_id: '1171281198'
frame_number: 20
v1_score: 0.317
- rank: 4
frame_id: '1171281208'
frame_number: 27
v1_score: 0.296
- rank: 5
frame_id: '1171281197'
frame_number: 19
v1_score: 0.23
v2_rerank:
- frame_id: '1171281190'
frame_number: 13
v1_rank: 1
v1_score: 0.563
semantic_score: 0.634
v2_rank: 1
- frame_id: '1171281192'
frame_number: 15
v1_rank: 2
v1_score: 0.517
semantic_score: 0.5959
v2_rank: 2
- frame_id: '1171281198'
frame_number: 20
v1_rank: 3
v1_score: 0.317
semantic_score: 0.5188
v2_rank: 3
- frame_id: '1171281208'
frame_number: 27
v1_rank: 4
v1_score: 0.296
semantic_score: 0.4577
v2_rank: 4
- frame_id: '1171281197'
frame_number: 19
v1_rank: 5
v1_score: 0.23
semantic_score: 0.4209
v2_rank: 5
03-2:
section_type: target
answer_frame_number: 29
mdx_title: 2. Process의 혁신과 Product의 변화
mdx_summary: 2. Process의 혁신과 Product의 변화. 과정 / 결과
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281210'
frame_number: 29
v1_score: 0.662
- rank: 2
frame_id: '1171281189'
frame_number: 12
v1_score: 0.523
- rank: 3
frame_id: '1171281182'
frame_number: 11
v1_score: 0.362
- rank: 4
frame_id: '1171281180'
frame_number: 9
v1_score: 0.358
- rank: 5
frame_id: '1171281179'
frame_number: 8
v1_score: 0.328
v2_rerank:
- frame_id: '1171281210'
frame_number: 29
v1_rank: 1
v1_score: 0.662
semantic_score: 0.5991
v2_rank: 1
- frame_id: '1171281179'
frame_number: 8
v1_rank: 5
v1_score: 0.328
semantic_score: 0.4832
v2_rank: 2
- frame_id: '1171281182'
frame_number: 11
v1_rank: 3
v1_score: 0.362
semantic_score: 0.465
v2_rank: 3
- frame_id: '1171281189'
frame_number: 12
v1_rank: 2
v1_score: 0.523
semantic_score: 0.4395
v2_rank: 4
- frame_id: '1171281180'
frame_number: 9
v1_rank: 4
v1_score: 0.358
semantic_score: 0.3975
v2_rank: 5
01-1:
section_type: holdout
answer_frame_number: null
mdx_title: 1. 용어 정의
mdx_summary: '1. 용어 정의. fontSize: ''0.8rem'',. 건설산업 / BIM / DX'
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281172'
frame_number: 1
v1_score: 0.62
- rank: 2
frame_id: '1171281208'
frame_number: 27
v1_score: 0.232
- rank: 3
frame_id: '1171281189'
frame_number: 12
v1_score: 0.21
- rank: 4
frame_id: '1171281182'
frame_number: 11
v1_score: 0.203
- rank: 5
frame_id: '1171281180'
frame_number: 9
v1_score: 0.198
v2_rerank:
- frame_id: '1171281180'
frame_number: 9
v1_rank: 5
v1_score: 0.198
semantic_score: 0.6017
v2_rank: 1
- frame_id: '1171281208'
frame_number: 27
v1_rank: 2
v1_score: 0.232
semantic_score: 0.5881
v2_rank: 2
- frame_id: '1171281182'
frame_number: 11
v1_rank: 4
v1_score: 0.203
semantic_score: 0.5507
v2_rank: 3
- frame_id: '1171281189'
frame_number: 12
v1_rank: 3
v1_score: 0.21
semantic_score: 0.547
v2_rank: 4
- frame_id: '1171281172'
frame_number: 1
v1_rank: 1
v1_score: 0.62
semantic_score: 0.5368
v2_rank: 5
02-1:
section_type: holdout
answer_frame_number: null
mdx_title: 1. DX의 궁극적 목표
mdx_summary: 1. DX의 궁극적 목표. ![DX의 궁극적 목표](/assets/images/궁극적목표.png). 안전과 품질 / 생산성 향상 / 소통과 신뢰
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281189'
frame_number: 12
v1_score: 0.731
- rank: 2
frame_id: '1171281191'
frame_number: 14
v1_score: 0.33
- rank: 3
frame_id: '1171281208'
frame_number: 27
v1_score: 0.273
- rank: 4
frame_id: '1171281174'
frame_number: 3
v1_score: 0.25
- rank: 5
frame_id: '1171281198'
frame_number: 20
v1_score: 0.192
v2_rerank:
- frame_id: '1171281198'
frame_number: 20
v1_rank: 5
v1_score: 0.192
semantic_score: 0.5895
v2_rank: 1
- frame_id: '1171281191'
frame_number: 14
v1_rank: 2
v1_score: 0.33
semantic_score: 0.5779
v2_rank: 2
- frame_id: '1171281189'
frame_number: 12
v1_rank: 1
v1_score: 0.731
semantic_score: 0.57
v2_rank: 3
- frame_id: '1171281208'
frame_number: 27
v1_rank: 3
v1_score: 0.273
semantic_score: 0.5467
v2_rank: 4
- frame_id: '1171281174'
frame_number: 3
v1_rank: 4
v1_score: 0.25
semantic_score: 0.4707
v2_rank: 5
02-2.1:
section_type: holdout
answer_frame_number: null
mdx_title: 2.1 업무 수행 과정(Process)의 변화
mdx_summary: 2.1 업무 수행 과정(Process)의 변화
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281182'
frame_number: 11
v1_score: 0.201
- rank: 2
frame_id: '1171281179'
frame_number: 8
v1_score: 0.168
- rank: 3
frame_id: '1171281174'
frame_number: 3
v1_score: 0.15
- rank: 4
frame_id: '1171281210'
frame_number: 29
v1_score: 0.141
- rank: 5
frame_id: '1171281180'
frame_number: 9
v1_score: 0.135
v2_rerank:
- frame_id: '1171281210'
frame_number: 29
v1_rank: 4
v1_score: 0.141
semantic_score: 0.5411
v2_rank: 1
- frame_id: '1171281179'
frame_number: 8
v1_rank: 2
v1_score: 0.168
semantic_score: 0.4897
v2_rank: 2
- frame_id: '1171281174'
frame_number: 3
v1_rank: 3
v1_score: 0.15
semantic_score: 0.4472
v2_rank: 3
- frame_id: '1171281182'
frame_number: 11
v1_rank: 1
v1_score: 0.201
semantic_score: 0.4092
v2_rank: 4
- frame_id: '1171281180'
frame_number: 9
v1_rank: 5
v1_score: 0.135
semantic_score: 0.3943
v2_rank: 5
04-1:
section_type: holdout
answer_frame_number: null
mdx_title: 1. DX에 대한 인식
mdx_summary: '1. DX에 대한 인식. width: ''100%'',. 기술 및 소프트웨어 이해도 / 효과와 효율성 / 인력 및 교육 / 경제적 부담 / 실무 및 적용성'
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281198'
frame_number: 20
v1_score: 0.234
- rank: 2
frame_id: '1171281172'
frame_number: 1
v1_score: 0.217
- rank: 3
frame_id: '1171281193'
frame_number: 16
v1_score: 0.196
- rank: 4
frame_id: '1171281179'
frame_number: 8
v1_score: 0.177
- rank: 5
frame_id: '1171281194'
frame_number: 17
v1_score: 0.171
v2_rerank:
- frame_id: '1171281198'
frame_number: 20
v1_rank: 1
v1_score: 0.234
semantic_score: 0.5321
v2_rank: 1
- frame_id: '1171281172'
frame_number: 1
v1_rank: 2
v1_score: 0.217
semantic_score: 0.5225
v2_rank: 2
- frame_id: '1171281194'
frame_number: 17
v1_rank: 5
v1_score: 0.171
semantic_score: 0.4473
v2_rank: 3
- frame_id: '1171281179'
frame_number: 8
v1_rank: 4
v1_score: 0.177
semantic_score: 0.4422
v2_rank: 4
- frame_id: '1171281193'
frame_number: 16
v1_rank: 3
v1_score: 0.196
semantic_score: 0.4299
v2_rank: 5
04-2.1:
section_type: holdout
answer_frame_number: null
mdx_title: 2.1 정책 및 발주 체계
mdx_summary: 2.1 정책 및 발주 체계. 실질적 기술 경쟁을 저해하는 정책 집행 / 적용 효과가 있는 사례도 없이 방침부터 도입 / 엔지니어링 S/W에 대한 개념 부재 / 기술투자
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281193'
frame_number: 16
v1_score: 0.373
- rank: 2
frame_id: '1171281175'
frame_number: 4
v1_score: 0.2
- rank: 3
frame_id: '1171281194'
frame_number: 17
v1_score: 0.171
- rank: 4
frame_id: '1171281190'
frame_number: 13
v1_score: 0.167
- rank: 5
frame_id: '1171281206'
frame_number: 26
v1_score: 0.165
v2_rerank:
- frame_id: '1171281194'
frame_number: 17
v1_rank: 3
v1_score: 0.171
semantic_score: 0.5077
v2_rank: 1
- frame_id: '1171281193'
frame_number: 16
v1_rank: 1
v1_score: 0.373
semantic_score: 0.4877
v2_rank: 2
- frame_id: '1171281175'
frame_number: 4
v1_rank: 2
v1_score: 0.2
semantic_score: 0.4578
v2_rank: 3
- frame_id: '1171281206'
frame_number: 26
v1_rank: 5
v1_score: 0.165
semantic_score: 0.4259
v2_rank: 4
- frame_id: '1171281190'
frame_number: 13
v1_rank: 4
v1_score: 0.167
semantic_score: 0.3247
v2_rank: 5
04-2.2:
section_type: holdout
answer_frame_number: null
mdx_title: 2.2 조직 및 수행 역량
mdx_summary: 2.2 조직 및 수행 역량. 공학적 개념 정립 부재 / ‘본업 기술력 확보’ 우선의 개념 부재 / DX/BIM의 근본 취지와 목표의 이해 부족 / 과거의 타성에 머무르고 있는 기술자 집단
top_k: 5
v1_top_k:
- rank: 1
frame_id: '1171281193'
frame_number: 16
v1_score: 0.501
- rank: 2
frame_id: '1171281190'
frame_number: 13
v1_score: 0.249
- rank: 3
frame_id: '1171281195'
frame_number: 18
v1_score: 0.223
- rank: 4
frame_id: '1171281194'
frame_number: 17
v1_score: 0.212
- rank: 5
frame_id: '1171281192'
frame_number: 15
v1_score: 0.204
v2_rerank:
- frame_id: '1171281194'
frame_number: 17
v1_rank: 4
v1_score: 0.212
semantic_score: 0.6408
v2_rank: 1
- frame_id: '1171281193'
frame_number: 16
v1_rank: 1
v1_score: 0.501
semantic_score: 0.5841
v2_rank: 2
- frame_id: '1171281192'
frame_number: 15
v1_rank: 5
v1_score: 0.204
semantic_score: 0.4636
v2_rank: 3
- frame_id: '1171281195'
frame_number: 18
v1_rank: 3
v1_score: 0.223
semantic_score: 0.4366
v2_rank: 4
- frame_id: '1171281190'
frame_number: 13
v1_rank: 2
v1_score: 0.249
semantic_score: 0.4128
v2_rank: 5