vault backup: 2026-01-02 16:10:47
This commit is contained in:
71
.obsidian/workspace.json
vendored
71
.obsidian/workspace.json
vendored
@@ -4,39 +4,24 @@
|
|||||||
"type": "split",
|
"type": "split",
|
||||||
"children": [
|
"children": [
|
||||||
{
|
{
|
||||||
"id": "14467186b06ff799",
|
"id": "0a337572ddbe0554",
|
||||||
"type": "tabs",
|
"type": "tabs",
|
||||||
"children": [
|
"children": [
|
||||||
{
|
{
|
||||||
"id": "d74ba463d3773b61",
|
"id": "92b19ec62110b284",
|
||||||
"type": "leaf",
|
"type": "leaf",
|
||||||
"state": {
|
"state": {
|
||||||
"type": "markdown",
|
"type": "markdown",
|
||||||
"state": {
|
"state": {
|
||||||
"file": "Tasks.md",
|
"file": "Themen/Recherche/Automated Essay Scoring.md",
|
||||||
"mode": "source",
|
"mode": "source",
|
||||||
"source": false
|
"source": false
|
||||||
},
|
},
|
||||||
"icon": "lucide-file",
|
"icon": "lucide-file",
|
||||||
"title": "Tasks"
|
"title": "Automated Essay Scoring"
|
||||||
}
|
|
||||||
},
|
|
||||||
{
|
|
||||||
"id": "1d18b9de81e07b7d",
|
|
||||||
"type": "leaf",
|
|
||||||
"state": {
|
|
||||||
"type": "markdown",
|
|
||||||
"state": {
|
|
||||||
"file": "Exposé/3_Verwendete_Methoden.md",
|
|
||||||
"mode": "source",
|
|
||||||
"source": false
|
|
||||||
},
|
|
||||||
"icon": "lucide-file",
|
|
||||||
"title": "3_Verwendete_Methoden"
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
],
|
]
|
||||||
"currentTab": 1
|
|
||||||
}
|
}
|
||||||
],
|
],
|
||||||
"direction": "vertical"
|
"direction": "vertical"
|
||||||
@@ -171,13 +156,13 @@
|
|||||||
"state": {
|
"state": {
|
||||||
"type": "outline",
|
"type": "outline",
|
||||||
"state": {
|
"state": {
|
||||||
"file": "Tasks.md",
|
"file": "Themen/Orga/Quellen Architektur.md",
|
||||||
"followCursor": false,
|
"followCursor": false,
|
||||||
"showSearch": false,
|
"showSearch": false,
|
||||||
"searchQuery": ""
|
"searchQuery": ""
|
||||||
},
|
},
|
||||||
"icon": "lucide-list",
|
"icon": "lucide-list",
|
||||||
"title": "Outline of Tasks"
|
"title": "Outline of Quellen Architektur"
|
||||||
}
|
}
|
||||||
},
|
},
|
||||||
{
|
{
|
||||||
@@ -209,45 +194,45 @@
|
|||||||
"obsidian-git:Open Git source control": false
|
"obsidian-git:Open Git source control": false
|
||||||
}
|
}
|
||||||
},
|
},
|
||||||
"active": "1d18b9de81e07b7d",
|
"active": "92b19ec62110b284",
|
||||||
"lastOpenFiles": [
|
"lastOpenFiles": [
|
||||||
|
"Themen/Recherche/Kontext Checking.md",
|
||||||
|
"Themen/Recherche/Automated Essay Scoring.md",
|
||||||
|
"Themen/Recherche/Natural Language Processing.md",
|
||||||
|
"Themen/Orga/User Stories.md",
|
||||||
|
"Themen/Orga/Epics.md",
|
||||||
|
"Themen/Recherche/Grammatik.md",
|
||||||
|
"Themen/assets/Architecture.png",
|
||||||
|
"Exposé/1_Ziele.md",
|
||||||
|
"Exposé/3_Verwendete_Methoden.md",
|
||||||
|
"Themen/Orga/Quellen Architektur.md",
|
||||||
|
"Themen/Orga/Architektur.md",
|
||||||
|
"Exposé/4_Zeitplan.md",
|
||||||
|
"Exposé/0_Problembeschreibung.md",
|
||||||
|
"Themen/Orga/Untitled",
|
||||||
|
"Umsetzung/1_Core_Layer/Explanation Engine.md",
|
||||||
|
"Themen/Recherche",
|
||||||
|
"Themen/Orga",
|
||||||
|
"Umsetzung/1_Core_Layer/Performance Metric Observator.md",
|
||||||
|
"Umsetzung/1_Core_Layer/Logger.md",
|
||||||
|
"0_Tools",
|
||||||
"Tasks.md",
|
"Tasks.md",
|
||||||
"Ideen Vorgehensweise.md",
|
"Ideen Vorgehensweise.md",
|
||||||
"Umsetzung/3_Service_Layer/Response Orchestrator.md",
|
"Umsetzung/3_Service_Layer/Response Orchestrator.md",
|
||||||
"Umsetzung/3_Service_Layer/Scoring Model.md",
|
"Umsetzung/3_Service_Layer/Scoring Model.md",
|
||||||
"Umsetzung/3_Service_Layer/API Endpoints.md",
|
"Umsetzung/3_Service_Layer/API Endpoints.md",
|
||||||
"Exposé/3_Verwendete_Methoden.md",
|
|
||||||
"Exposé/1_Ziele.md",
|
|
||||||
"Exposé/4_Zeitplan.md",
|
|
||||||
"Themen/Architektur.md",
|
|
||||||
"Umsetzung/1_Core_Layer/Plug-In-Manager.md",
|
"Umsetzung/1_Core_Layer/Plug-In-Manager.md",
|
||||||
"Umsetzung/3_Service_Layer/Socket Connector.md",
|
"Umsetzung/3_Service_Layer/Socket Connector.md",
|
||||||
"Umsetzung/2.2_Extended_Analysis_Plugins/Hallucination Checker.md",
|
"Umsetzung/2.2_Extended_Analysis_Plugins/Hallucination Checker.md",
|
||||||
"Umsetzung/2.2_Extended_Analysis_Plugins/Fact Checker.md",
|
"Umsetzung/2.2_Extended_Analysis_Plugins/Fact Checker.md",
|
||||||
"Umsetzung/2.2_Extended_Analysis_Plugins/Context Checker.md",
|
"Umsetzung/2.2_Extended_Analysis_Plugins/Context Checker.md",
|
||||||
"Umsetzung/2.1_Analysis_Orchestrator/Workflow.md",
|
"Umsetzung/2.1_Analysis_Orchestrator/Workflow.md",
|
||||||
"Umsetzung/2.1_Analysis_Orchestrator/Pipeline.md",
|
|
||||||
"Umsetzung/2_Core_Analysis/Repetition Checker.md",
|
|
||||||
"Umsetzung/2_Core_Analysis/Grammar Checker.md",
|
|
||||||
"Umsetzung/2_Core_Analysis/Spelling Checker.md",
|
|
||||||
"Umsetzung/2_Core_Analysis/Punctuation Checker.md",
|
|
||||||
"Umsetzung/1_Core_Layer/Documentation Tool.md",
|
|
||||||
"Umsetzung/1_Core_Layer/Performance Metric Observator.md",
|
|
||||||
"Umsetzung/1_Core_Layer/Port Listener.md",
|
|
||||||
"Umsetzung/1_Core_Layer/Logger.md",
|
|
||||||
"Umsetzung/1_Core_Layer/Parser.md",
|
|
||||||
"Umsetzung/1_Core_Layer/CLI.md",
|
|
||||||
"Umsetzung/3_Service_Layer",
|
"Umsetzung/3_Service_Layer",
|
||||||
"Umsetzung/2.2_Extended_Analysis_Plugins",
|
"Umsetzung/2.2_Extended_Analysis_Plugins",
|
||||||
"Umsetzung/2.1_Analysis_Orchestrator",
|
"Umsetzung/2.1_Analysis_Orchestrator",
|
||||||
"Umsetzung/2_Core_Analysis",
|
"Umsetzung/2_Core_Analysis",
|
||||||
"Umsetzung/1_Core_Layer",
|
"Umsetzung/1_Core_Layer",
|
||||||
"Themen/assets/Architecture.png",
|
|
||||||
"Themen/assets",
|
"Themen/assets",
|
||||||
"Untitled",
|
|
||||||
"Umsetzung",
|
|
||||||
"Untitled.base",
|
|
||||||
"Themen",
|
|
||||||
"Untitled.canvas"
|
"Untitled.canvas"
|
||||||
]
|
]
|
||||||
}
|
}
|
||||||
@@ -5,7 +5,7 @@ tags:
|
|||||||
Für das Erstellen der Thesis werden folgende Werkzeuge, die mit folgendem Zweck verwendet werden.
|
Für das Erstellen der Thesis werden folgende Werkzeuge, die mit folgendem Zweck verwendet werden.
|
||||||
|
|
||||||
| Werkzeug | Zweck |
|
| Werkzeug | Zweck |
|
||||||
| ---------------- | -------------------------------------------------------- |
|
| ---------------------------- | -------------------------------------------------------- |
|
||||||
| Obsidian | Knowledge Graph, Persönliche Dokumentation, Organisation |
|
| Obsidian | Knowledge Graph, Persönliche Dokumentation, Organisation |
|
||||||
| Gitea | Online Repository und Code Organisation, Management |
|
| Gitea | Online Repository und Code Organisation, Management |
|
||||||
| LaTeX / Typst | Schreiben der Thesis / Exposé |
|
| LaTeX / Typst | Schreiben der Thesis / Exposé |
|
||||||
@@ -14,7 +14,11 @@ Für das Erstellen der Thesis werden folgende Werkzeuge, die mit folgendem Zweck
|
|||||||
| VSCode | Development Environment |
|
| VSCode | Development Environment |
|
||||||
| Python | Ausgewählte Programmiersprache |
|
| Python | Ausgewählte Programmiersprache |
|
||||||
| Docker / Podman | Containerisierung |
|
| Docker / Podman | Containerisierung |
|
||||||
| CI/CD | Gitea Worker |
|
| Gitea Actions, Woodpecker CI | CI / CD |
|
||||||
|
| NGINX | Hosting der Dokumentation |
|
||||||
|
| Sphinx | Generierung der Dokumentation |
|
||||||
|
| pylint, yapf, isort | Code Linting |
|
||||||
|
| pre-commit | Linting Überprüfung vor Pushes auf Branches |
|
||||||
|
|
||||||
# Repository Strategien
|
# Repository Strategien
|
||||||
|
|
||||||
@@ -27,6 +31,35 @@ Für die CI/CD Pipeline sind folgende Module von Interesse:
|
|||||||
2. Modultests
|
2. Modultests
|
||||||
3. Automatische Generierung der Dokumentation
|
3. Automatische Generierung der Dokumentation
|
||||||
|
|
||||||
|
Für die CI/CD Pipeline wurde Woodpecker verwendet, dieses lässt sich in die bereits existierende Gitea Instanz hinzufügen. Zusätzlich ermöglicht dies eine Erweiterung der Integration und Development Pipelines im Sinne, dass von dort ein Automatisiertes Prüfen des Code-Stils und anschließend die Dokumentation erneuert werden kann.
|
||||||
# Code Stil
|
# Code Stil
|
||||||
Für den Code Stil wird sich an die Vorgaben durch Google gehalten, folgender [Styleguide](https://google.github.io/styleguide/pyguide.html) wird dafür verwendet. Dem Linter wird dieses Styleguide ebenfalls übergeben, sodass die Code-Architektur Einheitlich durchgeführt wird.
|
Für den Code Stil wird sich an die Vorgaben durch Google gehalten, folgender [Styleguide](https://google.github.io/styleguide/pyguide.html) wird dafür verwendet. Dem Linter wird dieses Styleguide ebenfalls übergeben, sodass die Code-Architektur Einheitlich durchgeführt wird.
|
||||||
|
|
||||||
|
Der Code-Stil wird durch mehrere Tools automatisiert begutachtet und beim nicht-einhalten kann kein Merging in den `main` Branch durchgeführt werden. Diese Tools sind folgende: *pre-commit*, *pylint*, *yapf*, *isort*
|
||||||
|
|
||||||
|
## Code Dokumentation
|
||||||
|
Zur Bereitstellung der Dokumentation wird [Sphinx](https://www.sphinx-doc.org/en/master/) verwendet.
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
# Dokumentation
|
||||||
|
Zusätzlich zur Code-Dokumentation wird in der Thesis die Dokumentation für die Programmnutzung als Anhang hinzugefügt werden.
|
||||||
|
|
||||||
|
|
||||||
|
# Architektur
|
||||||
|
Die Architektur wird anhand mehrere Prinzipien gestützt.
|
||||||
|
## Hybride Plug-in Layer + Pattern Architektur
|
||||||
|
|
||||||
|
## Analysis Orchestrator
|
||||||
|
_"Advantages include decoupling with logic layers, collating many functions, and acting as a centralized controller"_ [gaurgaurav+1](https://www.gaurgaurav.com/patterns/orchestration-pattern/)
|
||||||
|
|
||||||
|
Anhand des Orchestrators wird die Logikschicht und deren Plug-Ins orchestriert, sodass ein zentraler Controller für die Grundfunktionen und Erweiterungen existiert.
|
||||||
|
|
||||||
|
## Core Analysis
|
||||||
|
_"Predefined linguistic rules are used to analyze and process textual data... deterministic in nature, not probabilistic"_ [geeksforgeeks](https://www.geeksforgeeks.org/nlp/rule-based-approach-in-nlp/)
|
||||||
|
|
||||||
|
Dabei wird folgende Regel verfolgt:
|
||||||
|
_"A rule-based model is deterministic in nature... Either a document satisfies a given rule completely or it doesn't"_ [sas](https://support.sas.com/resources/papers/proceedings17/SAS0587-2017.pdf)
|
||||||
|
|
||||||
|
Die Umsetzung erfolgt Algorithmisch anstelle von mit KI wegen folgender Problematik:
|
||||||
|
_"Rule-based approach produces fewer minor hallucinations than neural counterparts"_ [aclanthology](https://aclanthology.org/2024.inlg-main.48.pdf)
|
||||||
264
Themen/Orga/Epics.md
Normal file
264
Themen/Orga/Epics.md
Normal file
@@ -0,0 +1,264 @@
|
|||||||
|
---
|
||||||
|
tags:
|
||||||
|
- Architektur
|
||||||
|
- Epic
|
||||||
|
- Aufgaben
|
||||||
|
- "#UserStories"
|
||||||
|
---
|
||||||
|
|
||||||
|
|
||||||
|
|Epic|Beschreibung (Outcome)|Abgedeckte Personas|
|
||||||
|
|---|---|---|
|
||||||
|
|**Textqualität & Sprachkorrektheit**|Texte sollen sprachlich korrekt, verständlich und formal sauber sein|Chris, Emily, Franzi|
|
||||||
|
|**Kontextuelle Textanalyse**|Texte sollen inhaltlich zusammenhängend und logisch verständlich sein|Max, Emily|
|
||||||
|
|**Wissenschaftliche & faktenbasierte Validierung**|Inhalte sollen faktenbasiert, überprüfbar und quellenbelegt sein|Sascha, Emily|
|
||||||
|
|**Transparente Fehler- & Quellenkennzeichnung**|Fehler und Quellen sollen klar sichtbar und nachvollziehbar dargestellt werden|Franzi, Sascha|
|
||||||
|
|**KI-Textanalyse & Qualitätsverbesserung**|KI-generierte Texte sollen analysiert und qualitativ verbessert werden|Miriam|
|
||||||
|
|**Performante & effiziente Analyse**|Analysen sollen schnell und zuverlässig durchgeführt werden|Franzi|
|
||||||
|
|**Erweiterbare Analyseplattform**|Das System soll modular erweiterbar und wartbar sein|Tobias (Dev), Alex (Admin)|
|
||||||
|
|**Erklärbarkeit & Nachvollziehbarkeit**|Analyseergebnisse sollen erklärbar und auditierbar sein|Laura, Sascha|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Überblick: Epics → Architekturbezug
|
||||||
|
|
||||||
|
Zur Orientierung zuerst die Zuordnung:
|
||||||
|
|
||||||
|
|Epic|Hauptbetroffene Architekturteile|
|
||||||
|
|---|---|
|
||||||
|
|Textqualität & Sprachkorrektheit|Core Analysis, Analysis Orchestrator|
|
||||||
|
|Kontextuelle Textanalyse|Extended Analysis Plug-Ins|
|
||||||
|
|Wissenschaftliche & faktenbasierte Validierung|Extended Analysis Plug-Ins|
|
||||||
|
|Transparente Fehler- & Quellenkennzeichnung|Service Layer|
|
||||||
|
|KI-Textanalyse & Qualitätsverbesserung|Core + Extended Analysis|
|
||||||
|
|Performante & effiziente Analyse|Core Layer, Orchestrator|
|
||||||
|
|Erweiterbare Analyseplattform|Core Layer (Plugin-Manager)|
|
||||||
|
|Erklärbarkeit & Nachvollziehbarkeit|Service Layer, Core|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 1: **Textqualität & Sprachkorrektheit**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Texte sollen sprachlich korrekt, konsistent und formal sauber sein.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Nutzer möchte ich Grammatikfehler automatisch erkennen lassen|
|
||||||
|
|Als Nutzer möchte ich Rechtschreibfehler erkennen lassen|
|
||||||
|
|Als Nutzer möchte ich Zeichensetzungsfehler erkennen lassen|
|
||||||
|
|Als Nutzer möchte ich Wiederholungen im Text erkennen lassen|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Core Analysis**
|
||||||
|
|
||||||
|
- Grammatik-Checker implementieren
|
||||||
|
- Rechtschreib-Checker implementieren
|
||||||
|
- Punkt- & Komma-Checker implementieren
|
||||||
|
- Repetition-Checker implementieren
|
||||||
|
|
||||||
|
**Analysis Orchestrator**
|
||||||
|
|
||||||
|
- Reihenfolge der Checks definieren
|
||||||
|
- Ergebnisse aggregieren
|
||||||
|
- Einheitliches Ergebnisformat erzeugen
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 2: **Kontextuelle Textanalyse**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Inhaltliche Zusammenhänge und logische Konsistenz prüfen.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Nutzer möchte ich erkennen, ob mein Text inhaltlich konsistent ist|
|
||||||
|
|Als Nutzer möchte ich Hinweise auf Kontextbrüche erhalten|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Extended Analysis Plug-Ins**
|
||||||
|
|
||||||
|
- Context-Checker-Plugin definieren
|
||||||
|
- Plugin-Interface implementieren
|
||||||
|
- Kontextbewertung (lokal / global)
|
||||||
|
|
||||||
|
**Analysis Orchestrator**
|
||||||
|
|
||||||
|
- Kontext-Check optional aktivierbar machen
|
||||||
|
- Abhängigkeit zu Core-Checks definieren
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 3: **Wissenschaftliche & faktenbasierte Validierung**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Texte sollen faktenbasiert, quellengeprüft und vertrauenswürdig sein.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Wissenschaftler möchte ich Fakten prüfen lassen|
|
||||||
|
|Als Nutzer möchte ich Aussagen mit Quellen belegt sehen|
|
||||||
|
|Als Nutzer möchte ich Halluzinationen erkennen|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Extended Analysis Plug-Ins**
|
||||||
|
|
||||||
|
- Fact-Checker-Plugin implementieren
|
||||||
|
- Hallucination-Checker-Plugin implementieren
|
||||||
|
- Externe Quellenanbindung definieren
|
||||||
|
|
||||||
|
**Core Layer**
|
||||||
|
|
||||||
|
- Plugin-Lifecycle (Load / Unload)
|
||||||
|
- Timeout-Handling für externe Checks
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 4: **Transparente Fehler- & Quellenkennzeichnung**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Analyseergebnisse sollen klar sichtbar und nachvollziehbar sein.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Nutzer möchte ich Fehler im Text markiert sehen|
|
||||||
|
|Als Nutzer möchte ich Quellen per Hover einsehen|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Service Layer**
|
||||||
|
|
||||||
|
- Response Orchestrator erweitern
|
||||||
|
- Markierungsformat definieren
|
||||||
|
- Hover-fähige Quellenstruktur erzeugen
|
||||||
|
|
||||||
|
**Core**
|
||||||
|
|
||||||
|
- Einheitliches Result-Model definieren
|
||||||
|
- Confidence / Severity berücksichtigen
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 5: **KI-Textanalyse & Qualitätsverbesserung**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
KI-Texte gezielt analysieren und verbessern.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Forscherin möchte ich Wiederholungen in KI-Texten messen|
|
||||||
|
|Als Forscherin möchte ich Analyseergebnisse exportieren|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Core Analysis**
|
||||||
|
|
||||||
|
- Wiederholungsmetriken definieren
|
||||||
|
|
||||||
|
**Service Layer**
|
||||||
|
|
||||||
|
- Exportformate (JSON)
|
||||||
|
- Analysehistorie ausgeben
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 6: **Performante & effiziente Analyse**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Analyse soll schnell, stabil und skalierbar sein.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Nutzer möchte ich schnelle Analyseergebnisse|
|
||||||
|
|Als Betreiber möchte ich Engpässe erkennen|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Core Layer**
|
||||||
|
|
||||||
|
- Performance Metric Observer implementieren
|
||||||
|
- Latenz pro Plugin messen
|
||||||
|
|
||||||
|
**Analysis Orchestrator**
|
||||||
|
|
||||||
|
- Parallele Plugin-Ausführung
|
||||||
|
- Abbruchlogik bei Timeouts
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 7: **Erweiterbare Analyseplattform**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Neue Analyse-Plug-ins sollen einfach integrierbar sein.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Entwickler möchte ich neue Plugins einfach einbinden|
|
||||||
|
|Als Admin möchte ich Plugins konfigurieren|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Core Layer**
|
||||||
|
|
||||||
|
- Plugin-Manager implementieren
|
||||||
|
- Plugin-Discovery
|
||||||
|
- Plugin-Konfiguration
|
||||||
|
|
||||||
|
**Dokumentation**
|
||||||
|
|
||||||
|
- Plugin-Interface dokumentieren
|
||||||
|
- Beispiel-Plugin bereitstellen
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
# Epic 8: **Erklärbarkeit & Nachvollziehbarkeit**
|
||||||
|
|
||||||
|
## Ziel
|
||||||
|
|
||||||
|
Analyseergebnisse müssen erklärbar und auditierbar sein.
|
||||||
|
|
||||||
|
### User Stories
|
||||||
|
|
||||||
|
|Story|
|
||||||
|
|---|
|
||||||
|
|Als Nutzer möchte ich verstehen, warum etwas als Fehler markiert wurde|
|
||||||
|
|Als Compliance-Beauftragte möchte ich Analyseentscheidungen nachvollziehen|
|
||||||
|
|
||||||
|
### Aufgaben (Tasks)
|
||||||
|
|
||||||
|
**Service Layer**
|
||||||
|
|
||||||
|
- Erklärungstexte generieren
|
||||||
|
- Quellen & Gründe anzeigen
|
||||||
|
|
||||||
|
**Core**
|
||||||
|
|
||||||
|
- Logging der Analyseentscheidungen
|
||||||
|
- Analyse-Audit-Trail
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
228
Themen/Orga/Quellen Architektur.md
Normal file
228
Themen/Orga/Quellen Architektur.md
Normal file
@@ -0,0 +1,228 @@
|
|||||||
|
Basierend auf meiner Analyse habe ich **11 wissenschaftliche Quellen** identifiziert, die verschiedene Aspekte Ihrer hybriden Plug-in- und Layer-Architektur unterstützen. Hier ist die Übersicht:
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## 🎯 **KERNUNTERSTÜTZUNG DER ARCHITEKTUR**
|
||||||
|
|
||||||
|
## **1. Hybrid Plug-in + Layer Architektur**
|
||||||
|
|
||||||
|
**✅ Plugin Architecture Pattern**
|
||||||
|
|
||||||
|
- **Quelle:** University of Waterloo CS446 [uwaterloo](https://cs.uwaterloo.ca/~m2nagapp/courses/CS446/1195/Arch_Design_Activity/PlugIn.pdf)
|
||||||
|
- **Unterstützt:** Core Layer mit Plugin-Manager, Extended Analysis Plug-Ins
|
||||||
|
- **Zitat:** _"Application logic is divided between independent plug-in modules and the basic core system, providing extensibility, flexibility, and isolation"_
|
||||||
|
|
||||||
|
|
||||||
|
**✅ Pure Plug-in Architectures**
|
||||||
|
- **Quelle:** ACM Queue (Birsan, 2005)[queue.acm](https://queue.acm.org/detail.cfm?id=1053345)
|
||||||
|
- **Unterstützt:** Plugin-Registry, Runtime-Management, Extension Points
|
||||||
|
- **Zitat:** _"Plug-ins themselves become hosts to other plug-ins by providing well-defined hook points (extension points)"_
|
||||||
|
|
||||||
|
|
||||||
|
**✅ Layered Architecture**
|
||||||
|
- **Quelle:** ScienceDirect Topics, Philadelphia.edu[philadelphia+1](https://www.philadelphia.edu.jo/academics/lalqoran/uploads/layeredarchitecture-120625120536-phpapp01.pdf)
|
||||||
|
- **Unterstützt:** Service Layer, Core Layer Trennung
|
||||||
|
- **Zitat:** _"Each layer has strong cohesion and only relies on lower layers"_
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## **2. Analysis Orchestrator (Pipeline/Workflow)**
|
||||||
|
|
||||||
|
**✅ Orchestration Workflows in Distributed Systems**
|
||||||
|
|
||||||
|
- **Quelle:** IJFMR (Singhal, 2024)[ijfmr](https://www.ijfmr.com/papers/2024/6/30191.pdf)
|
||||||
|
- **Unterstützt:** Analysis Orchestrator, Task Sequencing
|
||||||
|
- **Zitat:** _"Workflow definitions serve as the blueprint for service interactions... Task sequencing mechanisms ensure proper ordering"_
|
||||||
|
- **Empirisch:** 47% Reduktion von Integration-Fehlern durch Orchestration
|
||||||
|
|
||||||
|
|
||||||
|
**✅ Orchestrator Pattern**
|
||||||
|
- **Quellen:** Gaur Gaurav, Kislay Verma[gaurgaurav+1](https://www.gaurgaurav.com/patterns/orchestration-pattern/)
|
||||||
|
- **Unterstützt:** Zentraler Controller für Analyse-Schritte
|
||||||
|
- **Zitat:** _"Advantages include decoupling with logic layers, collating many functions, and acting as a centralized controller"_
|
||||||
|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## **3. Rule-Based Analysis (Core Analysis ohne KI)**
|
||||||
|
|
||||||
|
**✅ Rule-Based NLP Approach**
|
||||||
|
|
||||||
|
- **Quelle:** GeeksforGeeks[geeksforgeeks](https://www.geeksforgeeks.org/nlp/rule-based-approach-in-nlp/)
|
||||||
|
- **Unterstützt:** Grammar, Spelling, Punctuation Checker
|
||||||
|
- **Zitat:** _"Predefined linguistic rules are used to analyze and process textual data... deterministic in nature, not probabilistic"_
|
||||||
|
|
||||||
|
|
||||||
|
**✅ Best Practices in Rule-Based Models**
|
||||||
|
|
||||||
|
- **Quelle:** SAS[sas](https://support.sas.com/resources/papers/proceedings17/SAS0587-2017.pdf)
|
||||||
|
- **Unterstützt:** Algorithmen-basierte Textanalyse
|
||||||
|
- **Zitat:** _"A rule-based model is deterministic in nature... Either a document satisfies a given rule completely or it doesn't"_
|
||||||
|
|
||||||
|
**✅ Interpretable Rule-Based Systems**
|
||||||
|
|
||||||
|
- **Quelle:** ACL Anthology (Warczyński, 2024)[aclanthology](https://aclanthology.org/2024.inlg-main.48.pdf)
|
||||||
|
- **Unterstützt:** Regelbasierte Systeme ohne KI
|
||||||
|
- **Zitat:** _"Rule-based approach produces fewer minor hallucinations than neural counterparts"_
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## **4. Modular NLP Pipeline**
|
||||||
|
|
||||||
|
**✅ Modular Pipeline for NLP**
|
||||||
|
|
||||||
|
- **Quelle:** PMC[pmc.ncbi.nlm.nih](https://pmc.ncbi.nlm.nih.gov/articles/PMC12262768/)
|
||||||
|
- **Unterstützt:** Pipeline-Architektur, Modularität
|
||||||
|
- **Zitat:** _"Modular pipeline maximizes flexibility for future deployments... represents a practical middle ground"_
|
||||||
|
|
||||||
|
|
||||||
|
**✅ Scalable Architecture for Text Analysis**
|
||||||
|
- **Quelle:** Darmont et al. (ADMA 2016)[eric.univ-lyon2](https://eric.univ-lyon2.fr/jdarmont/publications/files/adma2016paper44-cr.pdf)
|
||||||
|
- **Unterstützt:** Skalierbare Textanalyse mit modularen Komponenten
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## **5. Explanation Engine (Epic 8)**
|
||||||
|
|
||||||
|
**✅ Four Principles of Explainable AI**
|
||||||
|
|
||||||
|
- **Quelle:** NIST IR 8312[nvlpubs.nist](https://nvlpubs.nist.gov/nistpubs/ir/2021/nist.ir.8312.pdf)
|
||||||
|
|
||||||
|
- **Unterstützt:** Explanation Engine im Core Layer
|
||||||
|
|
||||||
|
- **Zitat:** _"Systems must provide evidence, support, or reasoning for outputs... explanation must be accessible, understandable, and actionable"_
|
||||||
|
|
||||||
|
|
||||||
|
**✅ Cognitive Architecture for Explainable AI**
|
||||||
|
|
||||||
|
- **Quelle:** arXiv[arxiv](https://arxiv.org/html/2512.03072v1)
|
||||||
|
|
||||||
|
- **Unterstützt:** Audit Trail, Decision Tracing
|
||||||
|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## **6. Service Layer & Response Orchestrator**
|
||||||
|
|
||||||
|
**✅ Service Layer Pattern**
|
||||||
|
|
||||||
|
- **Quelle:** Martin Fowler (2003)[martinfowler](https://martinfowler.com/eaaCatalog/serviceLayer.html)
|
||||||
|
|
||||||
|
- **Unterstützt:** Service Layer Rolle
|
||||||
|
|
||||||
|
- **Zitat:** _"A Service Layer defines an application's boundary... coordinates the application's response in each operation"_
|
||||||
|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## **7. Separation of Concerns**
|
||||||
|
|
||||||
|
**✅ SoC Principle**
|
||||||
|
|
||||||
|
- **Quelle:** Dijkstra (1974), Wikipedia, GeeksforGeeks[wikipedia+1](https://en.wikipedia.org/wiki/Separation_of_concerns)
|
||||||
|
|
||||||
|
- **Unterstützt:** Layer-Trennung, Modularität
|
||||||
|
|
||||||
|
- **Zitat:** _"Module details can be addressed in isolation; module integration is treated as a separate concern"_
|
||||||
|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## 📊 **ARCHITEKTUR-KOMPONENTEN → WISSENSCHAFTLICHE QUELLEN**
|
||||||
|
|
||||||
|
|Komponente|Unterstützt durch|
|
||||||
|
|---|---|
|
||||||
|
|**Plugin-Manager (Core Layer)**|CS446[uwaterloo](https://cs.uwaterloo.ca/~m2nagapp/courses/CS446/1195/Arch_Design_Activity/PlugIn.pdf), ACM Queue[queue.acm](https://queue.acm.org/detail.cfm?id=1053345)|
|
||||||
|
|**Core Analysis (Grammar/Spelling/etc.)**|GeeksforGeeks[geeksforgeeks](https://www.geeksforgeeks.org/nlp/rule-based-approach-in-nlp/), SAS[sas](https://support.sas.com/resources/papers/proceedings17/SAS0587-2017.pdf), ACL[aclanthology](https://aclanthology.org/2024.inlg-main.48.pdf)|
|
||||||
|
|**Extended Analysis Plug-Ins**|CS446[uwaterloo](https://cs.uwaterloo.ca/~m2nagapp/courses/CS446/1195/Arch_Design_Activity/PlugIn.pdf), ACM Queue[queue.acm](https://queue.acm.org/detail.cfm?id=1053345)|
|
||||||
|
|**Analysis Orchestrator**|IJFMR[ijfmr](https://www.ijfmr.com/papers/2024/6/30191.pdf), Gaur[gaurgaurav](https://www.gaurgaurav.com/patterns/orchestration-pattern/), Verma[kislayverma](https://kislayverma.com/software-architecture/architecture-pattern-orchestration-via-workflows/)|
|
||||||
|
|**Response Orchestrator**|Martin Fowler[martinfowler](https://martinfowler.com/eaaCatalog/serviceLayer.html)|
|
||||||
|
|**Service Layer**|Fowler[martinfowler](https://martinfowler.com/eaaCatalog/serviceLayer.html), Philadelphia.edu[philadelphia](https://www.philadelphia.edu.jo/academics/lalqoran/uploads/layeredarchitecture-120625120536-phpapp01.pdf)|
|
||||||
|
|**Core Layer (Infrastruktur)**|Philadelphia[philadelphia](https://www.philadelphia.edu.jo/academics/lalqoran/uploads/layeredarchitecture-120625120536-phpapp01.pdf), Dijkstra[wikipedia](https://en.wikipedia.org/wiki/Separation_of_concerns)|
|
||||||
|
|**Explanation Engine**|NIST[nvlpubs.nist](https://nvlpubs.nist.gov/nistpubs/ir/2021/nist.ir.8312.pdf), arXiv[arxiv](https://arxiv.org/html/2512.03072v1)|
|
||||||
|
|**Performance Metric Observer**|IJFMR[ijfmr](https://www.ijfmr.com/papers/2024/6/30191.pdf)|
|
||||||
|
|**Modular Pipeline**|PMC[pmc.ncbi.nlm.nih](https://pmc.ncbi.nlm.nih.gov/articles/PMC12262768/), ADMA[eric.univ-lyon2](https://eric.univ-lyon2.fr/jdarmont/publications/files/adma2016paper44-cr.pdf)|
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## ✅ **FAZIT**
|
||||||
|
|
||||||
|
Ihre Architektur wird durch **11 akademische/wissenschaftliche Quellen** vollständig unterstützt:
|
||||||
|
|
||||||
|
1. **Hybrid Plug-in + Layer Pattern** → etabliert (CS446, ACM, ScienceDirect)
|
||||||
|
2. **Orchestrator Pattern** → dokumentiert (IJFMR, akademische Blogs)
|
||||||
|
3. **Rule-Based Analysis** → wissenschaftlich fundiert (ACL, SAS, GeeksforGeeks)
|
||||||
|
4. **Explainability** → NIST-Standard, Forschungsliteratur
|
||||||
|
5. **Service Layer** → Martin Fowler Pattern (Standardreferenz)
|
||||||
|
6. **Separation of Concerns** → Dijkstra (fundamentales Prinzip)
|
||||||
|
|
||||||
|
Die Architektur folgt etablierten Software-Engineering-Prinzipien und ist wissenschaftlich gut abgesichert. Die vollständige Dokumentation mit allen Zitaten habe ich in `scientific_sources_architecture.md` gespeichert.
|
||||||
|
|
||||||
|
1. [https://cs.uwaterloo.ca/~m2nagapp/courses/CS446/1195/Arch_Design_Activity/PlugIn.pdf](https://cs.uwaterloo.ca/~m2nagapp/courses/CS446/1195/Arch_Design_Activity/PlugIn.pdf)
|
||||||
|
2. [https://queue.acm.org/detail.cfm?id=1053345](https://queue.acm.org/detail.cfm?id=1053345)
|
||||||
|
3. [https://www.philadelphia.edu.jo/academics/lalqoran/uploads/layeredarchitecture-120625120536-phpapp01.pdf](https://www.philadelphia.edu.jo/academics/lalqoran/uploads/layeredarchitecture-120625120536-phpapp01.pdf)
|
||||||
|
4. [https://www.sciencedirect.com/topics/computer-science/layered-architecture](https://www.sciencedirect.com/topics/computer-science/layered-architecture)
|
||||||
|
5. [https://www.ijfmr.com/papers/2024/6/30191.pdf](https://www.ijfmr.com/papers/2024/6/30191.pdf)
|
||||||
|
6. [https://www.gaurgaurav.com/patterns/orchestration-pattern/](https://www.gaurgaurav.com/patterns/orchestration-pattern/)
|
||||||
|
7. [https://kislayverma.com/software-architecture/architecture-pattern-orchestration-via-workflows/](https://kislayverma.com/software-architecture/architecture-pattern-orchestration-via-workflows/)
|
||||||
|
8. [https://www.geeksforgeeks.org/nlp/rule-based-approach-in-nlp/](https://www.geeksforgeeks.org/nlp/rule-based-approach-in-nlp/)
|
||||||
|
9. [https://support.sas.com/resources/papers/proceedings17/SAS0587-2017.pdf](https://support.sas.com/resources/papers/proceedings17/SAS0587-2017.pdf)
|
||||||
|
10. [https://aclanthology.org/2024.inlg-main.48.pdf](https://aclanthology.org/2024.inlg-main.48.pdf)
|
||||||
|
11. [https://pmc.ncbi.nlm.nih.gov/articles/PMC12262768/](https://pmc.ncbi.nlm.nih.gov/articles/PMC12262768/)
|
||||||
|
12. [https://eric.univ-lyon2.fr/jdarmont/publications/files/adma2016paper44-cr.pdf](https://eric.univ-lyon2.fr/jdarmont/publications/files/adma2016paper44-cr.pdf)
|
||||||
|
13. [https://nvlpubs.nist.gov/nistpubs/ir/2021/nist.ir.8312.pdf](https://nvlpubs.nist.gov/nistpubs/ir/2021/nist.ir.8312.pdf)
|
||||||
|
14. [https://arxiv.org/html/2512.03072v1](https://arxiv.org/html/2512.03072v1)
|
||||||
|
15. [https://martinfowler.com/eaaCatalog/serviceLayer.html](https://martinfowler.com/eaaCatalog/serviceLayer.html)
|
||||||
|
16. [https://en.wikipedia.org/wiki/Separation_of_concerns](https://en.wikipedia.org/wiki/Separation_of_concerns)
|
||||||
|
17. [https://www.geeksforgeeks.org/software-engineering/separation-of-concerns-soc/](https://www.geeksforgeeks.org/software-engineering/separation-of-concerns-soc/)
|
||||||
|
18. [https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/attachments/images/156908260/cabdc5b3-14a7-401c-88df-a1824a00f495/Architecture.jpg](https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/attachments/images/156908260/cabdc5b3-14a7-401c-88df-a1824a00f495/Architecture.jpg)
|
||||||
|
19. [https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/attachments/images/156908260/15ee988f-2a62-457e-b8e3-7c7d60bbdf1f/Architecture.jpg](https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/attachments/images/156908260/15ee988f-2a62-457e-b8e3-7c7d60bbdf1f/Architecture.jpg)
|
||||||
|
20. [https://www.redhat.com/en/blog/hybrid-cloud-platform-layers](https://www.redhat.com/en/blog/hybrid-cloud-platform-layers)
|
||||||
|
21. [https://www.microsoft.com/en-us/research/publication/towards-modular-machine-learning-pipelines/](https://www.microsoft.com/en-us/research/publication/towards-modular-machine-learning-pipelines/)
|
||||||
|
22. [https://www.reddit.com/r/reactjs/comments/1hmk30p/how_to_design_a_pluginbased_architecture_in_react/](https://www.reddit.com/r/reactjs/comments/1hmk30p/how_to_design_a_pluginbased_architecture_in_react/)
|
||||||
|
23. [https://arxiv.org/pdf/2509.13487.pdf](https://arxiv.org/pdf/2509.13487.pdf)
|
||||||
|
24. [https://www.tencentcloud.com/techpedia/101584](https://www.tencentcloud.com/techpedia/101584)
|
||||||
|
25. [https://www.secoda.co/glossary/what-is-rule-based-classification](https://www.secoda.co/glossary/what-is-rule-based-classification)
|
||||||
|
26. [https://www.geeksforgeeks.org/nlp/natural-language-processing-nlp-pipeline/](https://www.geeksforgeeks.org/nlp/natural-language-processing-nlp-pipeline/)
|
||||||
|
27. [https://arxiv.org/abs/2510.04800](https://arxiv.org/abs/2510.04800)
|
||||||
|
28. [https://www.youtube.com/watch?v=wsPnnTMqEC4](https://www.youtube.com/watch?v=wsPnnTMqEC4)
|
||||||
|
29. [https://tnc-solutions.com/nlp-comment-fonctionne/](https://tnc-solutions.com/nlp-comment-fonctionne/)
|
||||||
|
30. [https://ionic.io/resources/articles/overcoming-the-challenges-of-mobile-app-architecture-with-hybrid-development](https://ionic.io/resources/articles/overcoming-the-challenges-of-mobile-app-architecture-with-hybrid-development)
|
||||||
|
31. [https://arxiv.org/html/2502.20609v1](https://arxiv.org/html/2502.20609v1)
|
||||||
|
32. [https://airbyte.com/data-engineering-resources/natural-language-processing-pipeline](https://airbyte.com/data-engineering-resources/natural-language-processing-pipeline)
|
||||||
|
33. [https://www.emergentmind.com/topics/hybrid-architecture-designs](https://www.emergentmind.com/topics/hybrid-architecture-designs)
|
||||||
|
34. [https://bookdown.org/valerie_hase/TextasData_HS2021/tutorial-12-rule-based-approaches-dictionaries.html](https://bookdown.org/valerie_hase/TextasData_HS2021/tutorial-12-rule-based-approaches-dictionaries.html)
|
||||||
|
35. [https://buildsimple.substack.com/p/extensibility-designing-for-future](https://buildsimple.substack.com/p/extensibility-designing-for-future)
|
||||||
|
36. [https://www.reddit.com/r/SoftwareEngineering/comments/rr8zp6/orchestrator_pattern_thoughts/](https://www.reddit.com/r/SoftwareEngineering/comments/rr8zp6/orchestrator_pattern_thoughts/)
|
||||||
|
37. [https://stackoverflow.com/questions/323202/how-to-design-extensible-software-plugin-architecture](https://stackoverflow.com/questions/323202/how-to-design-extensible-software-plugin-architecture)
|
||||||
|
38. [https://taylorandfrancis.com/knowledge/Engineering_and_technology/Computer_science/Separation_of_concerns/](https://taylorandfrancis.com/knowledge/Engineering_and_technology/Computer_science/Separation_of_concerns/)
|
||||||
|
39. [https://arxiv.org/abs/2510.00197](https://arxiv.org/abs/2510.00197)
|
||||||
|
40. [https://embeddedartistry.com/fieldmanual-terms/separation-of-concerns/](https://embeddedartistry.com/fieldmanual-terms/separation-of-concerns/)
|
||||||
|
41. [https://dl.acm.org/doi/full/10.1145/3698322.3698342](https://dl.acm.org/doi/full/10.1145/3698322.3698342)
|
||||||
|
42. [https://www.youtube.com/watch?v=JYb5pzcIimQ](https://www.youtube.com/watch?v=JYb5pzcIimQ)
|
||||||
|
43. [https://ris.utwente.nl/ws/files/5428452/Aksit01six.pdf](https://ris.utwente.nl/ws/files/5428452/Aksit01six.pdf)
|
||||||
|
44. [https://www.sciencedirect.com/science/article/abs/pii/S0045790616302026](https://www.sciencedirect.com/science/article/abs/pii/S0045790616302026)
|
||||||
|
45. [https://www.reddit.com/r/javahelp/comments/12wc0bf/building_an_extensible_plugin_system/](https://www.reddit.com/r/javahelp/comments/12wc0bf/building_an_extensible_plugin_system/)
|
||||||
|
46. [http://soft.vub.ac.be/Publications/2002/vub-prog-tr-02-16.pdf](http://soft.vub.ac.be/Publications/2002/vub-prog-tr-02-16.pdf)
|
||||||
|
47. [https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID4042351_code4193475.pdf?abstractid=4042351&mirid=1](https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID4042351_code4193475.pdf?abstractid=4042351&mirid=1)
|
||||||
|
48. [https://arobs.com/blog/plugin-based-architecture-in-modern-web-apps/](https://arobs.com/blog/plugin-based-architecture-in-modern-web-apps/)
|
||||||
|
49. [https://dev.to/tamerlan_dev/separation-of-concerns-the-simple-way-4jp2](https://dev.to/tamerlan_dev/separation-of-concerns-the-simple-way-4jp2)
|
||||||
|
50. [https://ink.library.smu.edu.sg/cgi/viewcontent.cgi?article=7580&context=sis_research](https://ink.library.smu.edu.sg/cgi/viewcontent.cgi?article=7580&context=sis_research)
|
||||||
|
51. [https://stackoverflow.com/questions/21339657/whats-the-difference-between-service-layer-and-domain-model-layer](https://stackoverflow.com/questions/21339657/whats-the-difference-between-service-layer-and-domain-model-layer)
|
||||||
|
52. [https://gerred.github.io/building-an-agentic-system/system-architecture-diagram.html](https://gerred.github.io/building-an-agentic-system/system-architecture-diagram.html)
|
||||||
|
53. [https://aclanthology.org/Y11-1026.pdf](https://aclanthology.org/Y11-1026.pdf)
|
||||||
|
54. [https://itnext.io/the-evolution-of-ddd-layered-architecture-6dd19b8dbc29](https://itnext.io/the-evolution-of-ddd-layered-architecture-6dd19b8dbc29)
|
||||||
|
55. [https://solace.com/event-driven-architecture-patterns/](https://solace.com/event-driven-architecture-patterns/)
|
||||||
|
56. [https://www.clevertype.co/post/the-science-behind-ai-grammar-correction](https://www.clevertype.co/post/the-science-behind-ai-grammar-correction)
|
||||||
|
57. [https://www.linkedin.com/pulse/layered-architecture-service-based-development-why-use-qicge](https://www.linkedin.com/pulse/layered-architecture-service-based-development-why-use-qicge)
|
||||||
|
58. [https://microservices.io/patterns/communication-style/messaging.html](https://microservices.io/patterns/communication-style/messaging.html)
|
||||||
|
59. [https://www.informatica.si/index.php/informatica/article/viewFile/368/369](https://www.informatica.si/index.php/informatica/article/viewFile/368/369)
|
||||||
|
60. [https://okenna.hashnode.dev/backend-communication-design-patterns-request-response-pattern](https://okenna.hashnode.dev/backend-communication-design-patterns-request-response-pattern)
|
||||||
|
61. [https://journal.esrgroups.org/jes/article/download/3611/2775/6494](https://journal.esrgroups.org/jes/article/download/3611/2775/6494)
|
||||||
|
62. [https://developer.android.com/topic/architecture/domain-layer](https://developer.android.com/topic/architecture/domain-layer)
|
||||||
|
63. [https://www.speakeasy.com/api-design/responses](https://www.speakeasy.com/api-design/responses)
|
||||||
|
64. [https://sparknlp.org/analyze_spelling_grammar](https://sparknlp.org/analyze_spelling_grammar)
|
||||||
|
65. [https://www.reddit.com/r/androiddev/comments/1c804jm/guide_to_app_architecture_what_is_the_domain_layer/](https://www.reddit.com/r/androiddev/comments/1c804jm/guide_to_app_architecture_what_is_the_domain_layer/)
|
||||||
|
66. [https://www.catchpoint.com/api-monitoring-tools/api-architecture](https://www.catchpoint.com/api-monitoring-tools/api-architecture)
|
||||||
|
67. [https://bagustris.github.io/nlp-python/09-gramatika/index.html](https://bagustris.github.io/nlp-python/09-gramatika/index.html)
|
||||||
|
68. [https://blog.stoplight.io/api-design-patterns-for-rest-web-services](https://blog.stoplight.io/api-design-patterns-for-rest-web-services)
|
||||||
35
Themen/Orga/User Stories.md
Normal file
35
Themen/Orga/User Stories.md
Normal file
@@ -0,0 +1,35 @@
|
|||||||
|
---
|
||||||
|
tags:
|
||||||
|
- UserStories
|
||||||
|
- Personas
|
||||||
|
---
|
||||||
|
|
||||||
|
# Quellen
|
||||||
|
[Atlassian - User Stories](https://www.atlassian.com/agile/project-management/user-stories)
|
||||||
|
|
||||||
|
# Methodik
|
||||||
|
Eine Vorversion wird anhand bereits existierenden wissen zur Erstellung mehrer Personas erstellt. Diese wird anschließend mit einer KI aufgebessert und final durch Eigenarbeit erneut überprüft. Die Personas sollen dabei alle Forderungen an das System darstellen, sodass das Erstellen von mehreren #Epic und spätere Bearbeitung und Planung in #OpenProject durchgeführt werden kann.
|
||||||
|
# Personas (Eigenarbeit)
|
||||||
|
|
||||||
|
| Persona | Forderungen | So what? |
|
||||||
|
| ------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------- |
|
||||||
|
| Emily | Ich möchte als Studentin, nachvollziehen können ob meine Texte wissenschaftlichen Standards entsprechen. | - Kontext prüfen<br>- Quellen prüfen |
|
||||||
|
| Sascha | Ich möchte als Wissenschaftler Quellenbasierte Texte von generativen KIs erhalten und diese bereits überprüft zurückbekommen. Informationen und Fehler sollen mit Quellen versehen sein, damit ich sicher damit arbeiten kann. | - Quellen prüfen<br>- Fehlermarkierungen<br>- Quellen durch Hovering kenntlich machen |
|
||||||
|
| Max | Als Schüler fällt es mir schwer den Kontext in meinen Texten richtig rüberbringen. Eine Software, die den Kontext für mich prüft wäre sehr schön. | - Kontext prüfen |
|
||||||
|
| Franzi | Als Arbeitnehmerin möchte ich schnelle und überprüfbare Texte von bereits existierenden KIs erhalten, die mit Markierungen versehen sind. Ich will mögliche Fehler bereits kenntlich gemacht haben um selber (mit einer anderen Software) Korrekturen vornehmen zu können. | - Performanz<br>- Quellen durch Hovering kenntlich machen<br>- Fehlermarkierungen |
|
||||||
|
| Chris | Als Schüler möchte ich meine Texte auf Grammatik, Rechtschreibung und Punkt und Komma-Setzung automatisch prüfen lassen. | - Grammatik prüfen<br>- Rechtschreibung prüfen<br>- Punkt-und-Kommasetzung prüfen |
|
||||||
|
| Miriam | Als Wissenschaftlerin im Bereich KI möchte ich überprüfen und KI so verbessern, dass diese sich weniger Wiederholt. Ein Tool, welches dies Automatisiert darstellt würde dies helfen. | - Wiederholungen prüfen |
|
||||||
|
|
||||||
|
# Personas (KI verbessert)
|
||||||
|
| Persona | Forderungen | So what? |
|
||||||
|
| ---------- | ----------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------- |
|
||||||
|
| Emily | Als Studentin möchte ich erkennen können, ob mein Text wissenschaftlichen Standards entspricht. | - Sicherheit bei Abgaben- Vermeidung formaler Fehler |
|
||||||
|
| Sascha | Als Wissenschaftler möchte ich KI-generierte Texte mit geprüften Quellen und markierten Fehlern erhalten. | - Vertrauenswürdige Inhalte- Nachvollziehbarkeit |
|
||||||
|
| Max | Als Schüler möchte ich Feedback zum inhaltlichen Zusammenhang meines Textes erhalten. | - Verständlichkeit verbessern |
|
||||||
|
| Franzi | Als Arbeitnehmerin möchte ich schnelle, überprüfbare Texte mit klaren Markierungen erhalten. | - Zeitersparnis- Effiziente Nachbearbeitung |
|
||||||
|
| Chris | Als Schüler möchte ich meine Texte automatisch auf Grammatik, Rechtschreibung und Zeichensetzung prüfen lassen. | - Sprachlich korrekte Texte |
|
||||||
|
| Miriam | Als KI-Forscherin möchte ich Wiederholungen in KI-Texten automatisch erkennen. | - Modellverbesserung |
|
||||||
|
| **Alex** | Als Systemadministrator möchte ich Plugins überwachen, konfigurieren und bei Fehlern eingreifen können. | - Stabiler Betrieb- Fehlerkontrolle |
|
||||||
|
| **Nina** | Als Produktmanagerin möchte ich Analyse-Ergebnisse aggregiert auswerten können. | - Produktverbesserung- Priorisierung von Features |
|
||||||
|
| **Tobias** | Als Entwickler möchte ich neue Analysis-Plugins einfach integrieren und testen können. | - Erweiterbarkeit- Wartbarkeit |
|
||||||
|
| **Laura** | Als Datenschutz-/Compliance-Beauftragte möchte ich nachvollziehen können, wie Analyse-Ergebnisse zustande kommen. | - Rechtssicherheit- Transparenz |
|
||||||
4
Themen/Recherche/Automated Essay Scoring.md
Normal file
4
Themen/Recherche/Automated Essay Scoring.md
Normal file
@@ -0,0 +1,4 @@
|
|||||||
|
Automated Essay Scoring scheint eine Teilgebiet zu sein, welches noch eine Menge arbeit erfordert und dementsprechend möglicherweise nicht all zu anfassbar zu sein scheint.
|
||||||
|
|
||||||
|
KI könnte hier zwar helfen, aber das hinzufügen einer Blackbox kann zu unerwünschten Nebeneffekte führen (schwere Nachvollziehbarkeit, mögliche Beeinflussung durch übergebene Texte).
|
||||||
|
|
||||||
87
Themen/Recherche/Kontext Checking.md
Normal file
87
Themen/Recherche/Kontext Checking.md
Normal file
@@ -0,0 +1,87 @@
|
|||||||
|
---
|
||||||
|
tags:
|
||||||
|
- Webscraping
|
||||||
|
- NLP
|
||||||
|
- FactChecking
|
||||||
|
- ContextChecking
|
||||||
|
- ExplanationGeneration
|
||||||
|
- AES
|
||||||
|
---
|
||||||
|
|
||||||
|
You can combine web scraping with NLP by building a small “fact‑checking” pipeline: extract a claim, scrape relevant pages, then use retrieval and natural‑language inference models to decide whether the web evidence supports or contradicts the claim.[aclanthology+1](https://aclanthology.org/2025.knowledgenlp-1.26/)
|
||||||
|
|
||||||
|
## Overall idea
|
||||||
|
|
||||||
|
- Turn the text you want to verify into one or more clear, atomic claims (e.g. “X happened on date Y”).[mbzuai+1](https://mbzuai.ac.ae/news/new-resources-for-fact-checking-llms-presented-at-emnlp/)
|
||||||
|
- Use web scraping or search APIs to collect pages from trusted sources related to the claim.
|
||||||
|
- Use NLP models to rank, filter, and compare those pages against the claim, then output a verdict such as supported / contradicted / not enough info.[sciencedirect+1](https://www.sciencedirect.com/science/article/abs/pii/S0952197625002842)
|
||||||
|
|
||||||
|
## Step 1: Extract and normalize claims
|
||||||
|
|
||||||
|
Use NLP to make the input “checkable”:
|
||||||
|
- Sentence splitting and claim extraction to break long text into short, single‑fact claims.[aclanthology+1](https://aclanthology.org/2025.emnlp-main.1615/)
|
||||||
|
- Optionally generate clarifying questions about the claim (who, when, where) to guide retrieval and disambiguation.[aclanthology+1](https://aclanthology.org/2025.knowledgenlp-1.26/)
|
||||||
|
|
||||||
|
Useful keywords/approaches: “atomic fact decomposition”, “checkworthy claim detection”, and “JEDI fact decomposition for NLI”.[github+1](https://github.com/Cartus/Automated-Fact-Checking-Resources)
|
||||||
|
|
||||||
|
## Step 2: Web scraping for evidence
|
||||||
|
|
||||||
|
Use web scraping or search to gather potential evidence:
|
||||||
|
|
||||||
|
- Formulate search queries from the claim (or the generated questions), then fetch top results from selected domains (news, Wikipedia, official sites). FIRE and other systems iteratively refine queries until they get good evidence.[aclanthology+1](https://aclanthology.org/2025.findings-naacl.158.pdf)
|
||||||
|
- Parse the pages with BeautifulSoup, Firecrawl, or similar; DEFAME is one example that combines automated scraping with fact‑checking.[edam+1](https://edam.org.tr/Uploads/Yukleme_Resim/pdf-28-08-2023-23-40-14.pdf)
|
||||||
|
|
||||||
|
Key points:
|
||||||
|
|
||||||
|
- Restrict to trusted / whitelisted domains to reduce noise.
|
||||||
|
- Store page title, URL, and cleaned text paragraphs for later scoring.
|
||||||
|
|
||||||
|
## Step 3: Retrieve and rank relevant snippets
|
||||||
|
|
||||||
|
Rather than feeding whole pages to the model, use retrieval to find the most relevant passages:
|
||||||
|
|
||||||
|
- Split pages into passages (e.g. 2–3 sentences) and compute embeddings (e.g. sentence‑transformers) to rank them by similarity to the claim.[acm+1](https://dl.acm.org/doi/10.1145/3477495.3531827)
|
||||||
|
- Generative retrieval (GERE) and dual‑stage BM25 + dense retrieval (as in Fathom) are standard patterns you can replicate.[aclanthology+1](https://aclanthology.org/anthology-files/pdf/fever/2025.fever-1.20.pdf)
|
||||||
|
|
||||||
|
This gives you a small set of top‑k evidence snippets that actually talk about the claim.
|
||||||
|
|
||||||
|
## Step 4: Compare claim vs evidence with NLP
|
||||||
|
|
||||||
|
Use Natural Language Inference (NLI) or fact‑checking models:
|
||||||
|
|
||||||
|
- Treat the claim as hypothesis and each evidence snippet as premise, and run an NLI model to classify each pair as supported / contradicted / unrelated.[arxiv+1](https://arxiv.org/html/2407.18367v1)
|
||||||
|
- Aggregate scores across snippets to produce a final label like supported, contradicted, or not enough info (NEI), similar to FEVER‑style fact‑checking.[arxiv+1](https://arxiv.org/abs/2110.14532)
|
||||||
|
|
||||||
|
Patterns you can adapt:
|
||||||
|
|
||||||
|
- VERITAS‑NLI: scrapes news articles and uses NLI models to verify headlines in real time.[sciencedirect](https://www.sciencedirect.com/science/article/abs/pii/S0952197625002842)
|
||||||
|
- ClaimCheck and FIRE: decompose claims, retrieve web evidence, then use smaller LMs/NLI models to derive a verdict.[aclanthology+1](https://aclanthology.org/2025.findings-naacl.158.pdf)
|
||||||
|
|
||||||
|
## Step 5: Produce explanation and handle uncertainty
|
||||||
|
|
||||||
|
For practical use, you should return more than just a label:
|
||||||
|
|
||||||
|
- Return a short natural‑language explanation plus top evidence snippets and URLs (e.g. “According to [source], X happened in 2022, not 2020”). Systems like FacTeR‑Check and explanation‑oriented fact‑checking pipelines use this pattern.[sciencedirect+1](https://www.sciencedirect.com/science/article/pii/S0952197624016506)
|
||||||
|
- If evidence is weak or conflicting, mark the result as “uncertain / needs human review” instead of forcing a yes/no answer.[arxiv+1](https://arxiv.org/html/2412.15189v3)
|
||||||
|
|
||||||
|
|
||||||
|
1. [https://aclanthology.org/2025.knowledgenlp-1.26/](https://aclanthology.org/2025.knowledgenlp-1.26/)
|
||||||
|
2. [https://www.sciencedirect.com/science/article/abs/pii/S0952197625002842](https://www.sciencedirect.com/science/article/abs/pii/S0952197625002842)
|
||||||
|
3. [https://mbzuai.ac.ae/news/new-resources-for-fact-checking-llms-presented-at-emnlp/](https://mbzuai.ac.ae/news/new-resources-for-fact-checking-llms-presented-at-emnlp/)
|
||||||
|
4. [https://aclanthology.org/2025.emnlp-main.1615/](https://aclanthology.org/2025.emnlp-main.1615/)
|
||||||
|
5. [https://github.com/Cartus/Automated-Fact-Checking-Resources](https://github.com/Cartus/Automated-Fact-Checking-Resources)
|
||||||
|
6. [https://aclanthology.org/2025.findings-naacl.158.pdf](https://aclanthology.org/2025.findings-naacl.158.pdf)
|
||||||
|
7. [https://aclanthology.org/anthology-files/pdf/fever/2025.fever-1.20.pdf](https://aclanthology.org/anthology-files/pdf/fever/2025.fever-1.20.pdf)
|
||||||
|
8. [https://edam.org.tr/Uploads/Yukleme_Resim/pdf-28-08-2023-23-40-14.pdf](https://edam.org.tr/Uploads/Yukleme_Resim/pdf-28-08-2023-23-40-14.pdf)
|
||||||
|
9. [https://github.com/multimodal-ai-lab/DEFAME](https://github.com/multimodal-ai-lab/DEFAME)
|
||||||
|
10. [https://dl.acm.org/doi/10.1145/3477495.3531827](https://dl.acm.org/doi/10.1145/3477495.3531827)
|
||||||
|
11. [https://arxiv.org/html/2407.18367v1](https://arxiv.org/html/2407.18367v1)
|
||||||
|
12. [https://arxiv.org/abs/2110.14532](https://arxiv.org/abs/2110.14532)
|
||||||
|
13. [https://www.sciencedirect.com/science/article/pii/S0952197624016506](https://www.sciencedirect.com/science/article/pii/S0952197624016506)
|
||||||
|
14. [https://arxiv.org/html/2412.15189v3](https://arxiv.org/html/2412.15189v3)
|
||||||
|
15. [https://www.biz4group.com/blog/developing-ai-automated-fact-checking-system](https://www.biz4group.com/blog/developing-ai-automated-fact-checking-system)
|
||||||
|
16. [https://download.hrz.tu-darmstadt.de/pub/FB20/Dekanat/Publikationen/AIPHES/Andreas_Hanselowski_NIPS-WPOC-2017.pdf](https://download.hrz.tu-darmstadt.de/pub/FB20/Dekanat/Publikationen/AIPHES/Andreas_Hanselowski_NIPS-WPOC-2017.pdf)
|
||||||
|
17. [https://journals.itb.ac.id/index.php/jictra/article/view/24157](https://journals.itb.ac.id/index.php/jictra/article/view/24157)
|
||||||
|
18. [https://search.gesis.org/research_data/SDN-10.7802-2469](https://search.gesis.org/research_data/SDN-10.7802-2469)
|
||||||
|
19. [https://arxiv.org/abs/2504.18376](https://arxiv.org/abs/2504.18376)
|
||||||
|
20. [https://ui.adsabs.harvard.edu/abs/arXiv:2409.00061](https://ui.adsabs.harvard.edu/abs/arXiv:2409.00061)
|
||||||
Binary file not shown.
|
Before Width: | Height: | Size: 112 KiB After Width: | Height: | Size: 112 KiB |
0
Umsetzung/1_Core_Layer/Explanation Engine.md
Normal file
0
Umsetzung/1_Core_Layer/Explanation Engine.md
Normal file
@@ -0,0 +1,19 @@
|
|||||||
|
Das Dokumentieren der Performanz ist von Interesse für weitere Iterationen. Hiermit kann in Zukunft folgendes geplant werden:
|
||||||
|
1. Performance Patches / Optimierungen
|
||||||
|
2. Issue-Detection
|
||||||
|
3. Scalability proaktiv unterstützen
|
||||||
|
4. Fehlerrate
|
||||||
|
|
||||||
|
# Anwendungsorientierung
|
||||||
|
Von Interesse sind #continuousMonitoring und #syntheticMonitoring aufgrund
|
||||||
|
folgender Kriterien:
|
||||||
|
|
||||||
|
- mimic real life scenarios
|
||||||
|
- key metric monitoring
|
||||||
|
- real-time observation
|
||||||
|
# Performance Baseline
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
# Sources
|
||||||
|
[Middleware-python-performance-monitoring](https://middleware.io/blog/python-performance-monitoring/)
|
||||||
|
|||||||
Reference in New Issue
Block a user