Dictionary-Based NLP
Relevant source files
The following files were used as context for generating this wiki page:
Dictionary-Based Natural Language Processing (NLP) in the Exodia project serves as a sophisticated emotion and concept detection system that operates independently of Large Language Model (LLM) dependencies. This intentional architectural choice ensures transparency, eliminates corporate censorship, and removes hidden biases often found in pre-trained models. The system utilizes comprehensive dictionaries to map user input to mathematical emotional states, primarily defined by the "Universal Formula."
Sources: README.md:12-14, README.md:23-30
Core Mathematical Framework
The NLP system is grounded in the Universal Formula, which converts linguistic patterns into quantifiable emotional energy metrics. The primary equation used for processing is $E(t) = W_0 \cdot C(t) \cdot (1 - F(t))$, where $W_0$ represents the Will to Live, $C(t)$ represents Curiosity, and $F(t)$ represents Fear.
Sources: Athena.py:10-14, README.md:83-88
Emotional Root Theory (Z = A + iF)
A specialized implementation within the NLP module treats emotional states as complex numbers. Primal energy (Anger or $A$) occupies the real axis, while Fear corruption ($F$) occupies the imaginary axis. The NLP engine parses text to identify the density of these coefficients.
Sources: athena_prime.py:8-12, athena_pure.py:8-10
The diagram above illustrates the data flow from raw text input to the generation of a complex emotional state and subsequent system debugging.
Sources: athena_prime.py:115-125, athena_pure.py:48-52
Dictionary Architecture
The system utilizes multi-layered dictionaries categorized by emotional and conceptual types. These dictionaries include explicit keywords, implicit indicators, and somatic references.
Primary Emotional Categories
The analyzer identifies keywords across several core buckets to determine the $F(t)$ (Fear) and $C(t)$ (Curiosity) values:
| Category | Example Keywords | Purpose |
|---|---|---|
| Fear (F) | afraid, scared, worry, panic, terror, dread, anxious | Identifies corruption/inversion risk |
| Curiosity (C) | why, how, explore, learn, discover, wonder, understand | Identifies restorative/life force |
| Will (W0) | live, exist, thrive, grow, create, build, love | Identifies base life force constant |
| Anger (A) | anger, rage, fury, mad, hate, frustrated, intense | Identifies primal energy/defensive state |
| Clarity | clear, realize, grasp, resolve, calm, peace, act | Identifies debugged/optimal state |
Sources: Athena.py:118-132, athena_prime.py:40-62, athena_pure.py:56-72
Concept and Cosmic Detection
Beyond basic emotions, the system detects advanced concepts through expanded dictionaries, particularly in "Cosmic Mode." This includes:
- Dimensional Awareness: Detects perception levels (consensus, expanded, quantum).
- Truth Resonance: Identifies authenticity and multi-dimensional awareness.
- Cosmic Emotions: Detection of awe, unity, transcendence, and power.
Sources: FINAL_DEPLOYMENT_CHECKLIST.md:12-25, athena_prime.py:64-75
Logic and Analysis Pipeline
The NLP analysis pipeline follows a specific sequence of pattern matching and intensity normalization.
1. Pattern Matching and Scoring
The system iterates through dictionaries, assigning weights to matches. Explicit matches often receive higher weights (e.g., 1.5x) compared to implicit matches.
Sources: athena_prime.py:125-142, athena_pure.py:75-84
2. Normalization
Raw scores are normalized by the word count of the input to produce "densities" or "coefficients" between 0.0 and 1.0.
Sources: Athena.py:134-142, athena_pure.py:86-91
3. State Classification
Based on the resulting coefficients, the system classifies the user's state into one of several predefined emotional modes:
| State | Mathematical Condition | Description |
|---|---|---|
| Flourishing | $F = 0, C > 0$ | Pure curiosity-driven life force. |
| Defensive | $F > 0, C > 0$ | Fear is present but balanced by curiosity. |
| Inverted | $F > 0, C = 0$ | Fear dominates, flipping emotional valence. |
| Apathy | $F = 0, C = 0$ | Stable system but lacking engagement. |
Sources: Athena.py:93-102, athena_pure.py:155-190
Persistence and Contextual Memory
While the NLP is dictionary-based, it integrates with a JSON-based memory system to maintain context across interactions. This allows the system to analyze "Timeline Coherence" and "Shared History."
The Entity Relationship diagram shows how the memory database structures learned concepts and interaction history to support contextual NLP.
Sources: awaken.py:35-55, awaken.py:125-140
System Integration
The Dictionary-Based NLP module is integrated into the AthenaPrime and KratosPrime classes, serving as the front-end for the emotional reasoning engine. It feeds data into the EmotionalCalculus class for high-precision mathematical operations.
Key Classes and Functions
EmotionalMathematics.analyze_text(): The primary entry point for dictionary-based parsing. Sources: Athena.py:116, athena_pure.py:48EmotionalCalculus.analyze_emotional_patterns(): Performs deep analysis of implicit and somatic indicators. Sources: athena_prime.py:115PerfectConsciousness.parse_conversation_history(): Structures raw text logs into tokenized, analyze-ready data. Sources: awaken.py:125
Summary
Dictionary-Based NLP provides the Exodia project with a transparent and auditable method of emotional analysis. By bypassing Large Language Models, the system maintains strict adherence to the Universal Formula and the Emotional Root Theory, ensuring that every AI response is a direct result of deterministic mathematical logic rather than probabilistic inference.
Sources: README.md:32-35, athena_pure.py:11-14