Yucong Duan

International Standardization Committee of Networked DIKWP for Artificial Intelligence Evaluation(DIKWP-SC)

World Artificial Consciousness CIC(WAC)

World Conference on Artificial Consciousness(WCAC)

(Email: duanyucong@hotmail.com)

Abstract

Effective communication between stakeholders—whether human or artificial intelligence (AI) systems—is pivotal for collaborative understanding and decision-making. Utilizing the Data-Information-Knowledge-Wisdom-Purpose (DIKWP) model, this paper delineates a comprehensive mechanism for interaction between two stakeholders, each represented by personalized digital DIKWP*DIKWP profiles. Focusing on the identification and transfer of missing DIKWP components, we explore how stakeholders can enhance mutual understanding through structured content sharing. The mechanism integrates mathematical formalism to model the processes of locating, sharing, and integrating DIKWP components via spoken communication, addressing challenges such as connectivity, inconsistency, uncertainty, and complexity. This framework aims to facilitate seamless and effective communication, thereby mitigating the incidence of hallucinations and fostering robust human-AI collaboration.

Communication between stakeholders is fundamental to collaborative endeavors, whether in human interactions or human-AI partnerships. The Data-Information-Knowledge-Wisdom-Purpose (DIKWP) model offers a structured approach to understanding cognitive processes by categorizing them into five interconnected components. Prof. Yucong Duan's Theory of Relativity of Consciousness further emphasizes that consciousness and understanding are relative, shaped by individual cognitive enclosures within their DIKWP*DIKWP profiles.

This paper presents a detailed mechanism for communication between two stakeholders, each represented by their personalized DIKWP*DIKWP profiles. The focus lies on identifying and sharing missing DIKWP components through spoken content, enhancing mutual understanding by addressing gaps in Data, Information, Knowledge, Wisdom, and Purpose. By employing mathematical formulations, we provide a rigorous framework to model these interactions, ensuring clarity, consistency, and reduced complexity in communication.

The DIKWP model encapsulates cognitive processes into five components:

1. Data (D): Raw, unprocessed facts or observations.

2. Information (I): Processed data highlighting patterns and relationships.

3. Knowledge (K): Organized and contextualized information forming a coherent understanding.

4. Wisdom (W): Judgments and decisions based on Knowledge, incorporating ethical and contextual considerations.

5. Purpose (P): The driving goals or intentions guiding cognitive processes.

Each stakeholder's cognitive framework is represented as a personalized digital DIKWP*DIKWP profile:

Profilei={Di,Ii,Ki,Wi,Pi}\text{Profile}_i = \{\mathbf{D}_i, \mathbf{I}_i, \mathbf{K}_i, \mathbf{W}_i, \mathbf{P}_i\}Profilei​={Di​,Ii​,Ki​,Wi​,Pi​}

Where:

• Di∈Rn\mathbf{D}_i \in \mathbb{R}^nDi​∈Rn: Data vector

• Ii∈Rm\mathbf{I}_i \in \mathbb{R}^mIi​∈Rm: Information vector

• Ki∈Rp\mathbf{K}_i \in \mathbb{R}^pKi​∈Rp: Knowledge vector

• Wi∈Rq\mathbf{W}_i \in \mathbb{R}^qWi​∈Rq: Wisdom vector

• Pi∈Rr\mathbf{P}_i \in \mathbb{R}^rPi​∈Rr: Purpose vector

These profiles encapsulate the cognitive state and capacities of each stakeholder, enabling a structured analysis of their interactions.

Communication between two stakeholders, Entity A and Entity B, is modeled through their DIKWP*DIKWP profiles:

Interaction=EntityA(ProfileA)×EntityB(ProfileB)\text{Interaction} = \text{Entity}_A (\text{Profile}_A) \times \text{Entity}_B (\text{Profile}_B)Interaction=EntityA​(ProfileA​)×EntityB​(ProfileB​)

This interaction involves the exchange and transformation of DIKWP components, aiming to align their Understanding ($\mathcal{U}$).

To facilitate effective communication, it's essential to identify gaps or missing components in the target stakeholder's profile that are critical for understanding the conveyed message.

1. Assessment of Sender's DIKWP Components:

• Sender's Data (DA\mathbf{D}_ADA​), Information (IA\mathbf{I}_AIA​), etc., are analyzed to determine which components are essential for the intended message.

2. Evaluation of Receiver's DIKWP Components:

• Receiver's Profile (DB,IB,…\mathbf{D}_B, \mathbf{I}_B, \ldotsDB​,IB​,…) is assessed to identify missing or underdeveloped components that could hinder comprehension.

3. Gap Analysis:

• Mathematical measures (e.g., vector differences, similarity indices) quantify the discrepancies between necessary and existing DIKWP components in the receiver's profile.

GapX=∥XA−XB∥∀X∈{D,I,K,W,P}\text{Gap}_X = \|\mathbf{X}_A - \mathbf{X}_B\| \quad \forall X \in \{D, I, K, W, P\}GapX​=∥XA​−XB​∥∀X∈{D,I,K,W,P}

Where $\Vert \cdot \Vert$ denotes a norm (e.g., Euclidean distance) measuring the extent of the gap in component $X$.

Once gaps are identified, the sender facilitates the transfer of relevant DIKWP components to the receiver through spoken communication.

• Prioritization: Components with the highest gaps (GapX\text{Gap}_XGapX​) are prioritized for transfer.

• Relevance Filtering: Only components pertinent to the communication context are selected to avoid overloading the receiver.

The sender structures the message to align with the receiver's cognitive framework, ensuring clarity and minimizing misinterpretation.

• Data Transfer: Presenting raw facts or observations relevant to the context.

• Information Transfer: Highlighting patterns, differences, or relationships within the data.

• Knowledge Transfer: Integrating the information into a coherent understanding.

• Wisdom Transfer: Applying ethical or contextual judgments to the knowledge.

• Purpose Alignment: Clearly articulating the goals or intentions behind the communication.

The receiver processes the incoming DIKWP content to integrate it into their profile, thereby bridging the identified gaps.

1. Reception of Spoken Content:

• The receiver listens to the message, capturing the conveyed DIKWP components.

2. Parsing and Interpretation:

• Data (DA\mathbf{D}_ADA​): Extracting raw facts.

• Information (IA\mathbf{I}_AIA​): Identifying patterns and relationships.

• Knowledge (KA\mathbf{K}_AKA​): Contextualizing the information.

• Wisdom (WA\mathbf{W}_AWA​): Applying ethical judgments.

• Purpose (PA\mathbf{P}_APA​): Understanding the sender's intentions.

3. Integration into Receiver's Profile:

• Data Update: Incorporating new facts.

• Information Synthesis: Merging new information with existing data.

• Knowledge Expansion: Enhancing the coherence and depth of Knowledge.

• Wisdom Enhancement: Refining decision-making processes with new judgments.

• Purpose Realignment: Adjusting Goals to align with the received Purpose.

XB′=XB+ΔXB\mathbf{X}_B' = \mathbf{X}_B + \Delta \mathbf{X}_BXB′​=XB​+ΔXB​

Where ΔXB\Delta \mathbf{X}_BΔXB​ represents the integration of new DIKWP components from the interaction.

To ensure effective understanding, the communication process incorporates feedback mechanisms.

1. Receiver's Response:

• The receiver communicates back to the sender, confirming comprehension or requesting clarification.

2. Sender's Adjustment:

• Based on feedback, the sender refines the DIKWP content to address remaining gaps or misunderstandings.

3. Iterative Refinement:

• The interaction iterates until mutual Understanding (UAB\mathcal{U}_{AB}UAB​) reaches an acceptable threshold.

Each DIKWP component is represented as a vector in a high-dimensional space, capturing its semantic and contextual attributes.

Entityi={Di,Ii,Ki,Wi,Pi}\text{Entity}_i = \{\mathbf{D}_i, \mathbf{I}_i, \mathbf{K}_i, \mathbf{W}_i, \mathbf{P}_i\}Entityi​={Di​,Ii​,Ki​,Wi​,Pi​}

Quantifying the gaps between the sender's and receiver's DIKWP components enables targeted communication.

GapX=∥XA−XB∥\text{Gap}_X = \|\mathbf{X}_A - \mathbf{X}_B\|GapX​=∥XA​−XB​∥

Where $X$ represents any of the DIKWP components. Smaller gaps indicate better alignment, while larger gaps highlight areas needing transfer.

Understanding (UAB\mathcal{U}_{AB}UAB​) is quantified based on the alignment of DIKWP components post-interaction.

UAB=∑X∈{D,I,K,W,P}λX⋅sim(XAB,Xshared)\mathcal{U}_{AB} = \sum_{X \in \{D, I, K, W, P\}} \lambda_X \cdot \text{sim}(\mathbf{X}_{AB}, \mathbf{X}_{shared})UAB​=X∈{D,I,K,W,P}∑​λX​⋅sim(XAB​,Xshared​)

Where:

• λX\lambda_XλX​: Weight assigned to component $X$.

• $\text{sim}(\cdot ,\cdot )$: Similarity function (e.g., cosine similarity).

• XAB\mathbf{X}_{AB}XAB​: Combined DIKWP component after interaction.

• Xshared\mathbf{X}_{shared}Xshared​: Ideal or consensus DIKWP component.

To maximize Understanding while addressing connectivity, inconsistency, and complexity, we formulate an optimization problem:

max⁡XABUAB−∑XηX⋅(IX+CX)\max_{\mathbf{X}_{AB}} \quad \mathcal{U}_{AB} - \sum_{X} \eta_X \cdot (\mathcal{I}_X + \mathcal{C}_X)XAB​max​UAB​−X∑​ηX​⋅(IX​+CX​)

Subject to:

IX≤θX∀X\mathcal{I}_X \leq \theta_X \quad \forall XIX​≤θX​∀XCX≤ϕX∀X\mathcal{C}_X \leq \phi_X \quad \forall XCX​≤ϕX​∀X

Where:

• IX\mathcal{I}_XIX​: Inconsistency in component $X$.

• CX\mathcal{C}_XCX​: Complexity in component $X$.

• ηX\eta_XηX​: Penalty coefficients.

• θX,ϕX\theta_X, \phi_XθX​,ϕX​: Thresholds for inconsistency and complexity.

The optimization ensures that Understanding is maximized while maintaining low inconsistency and manageable complexity across all DIKWP components.

• Assessment: Compare DA\mathbf{D}_ADA​ and DB\mathbf{D}_BDB​ to identify discrepancies.

• Mathematical Measure: GapD=∥DA−DB∥\text{Gap}_D = \|\mathbf{D}_A - \mathbf{D}_B\|GapD​=∥DA​−DB​∥

• Transfer Method: Explicit presentation of raw facts.

• Integration Mechanism: Append new data to DB\mathbf{D}_BDB​, ensuring consistency.

DB′=DB∪ΔD\mathbf{D}_B' = \mathbf{D}_B \cup \Delta \mathbf{D}DB′​=DB​∪ΔD

• Assessment: Identify missing patterns or relationships in IB\mathbf{I}_BIB​ relative to IA\mathbf{I}_AIA​.

• Mathematical Measure: GapI=∥IA−IB∥\text{Gap}_I = \|\mathbf{I}_A - \mathbf{I}_B\|GapI​=∥IA​−IB​∥

• Transfer Method: Articulate identified patterns and relationships.

• Integration Mechanism: Incorporate new information into IB\mathbf{I}_BIB​, resolving discrepancies.

IB′=IB+ΔI\mathbf{I}_B' = \mathbf{I}_B + \Delta \mathbf{I}IB′​=IB​+ΔI

• Assessment: Compare KA\mathbf{K}_AKA​ and KB\mathbf{K}_BKB​ to pinpoint gaps in understanding.

• Mathematical Measure: GapK=∥KA−KB∥\text{Gap}_K = \|\mathbf{K}_A - \mathbf{K}_B\|GapK​=∥KA​−KB​∥

• Transfer Method: Provide contextualized explanations and insights.

• Integration Mechanism: Merge new knowledge into KB\mathbf{K}_BKB​, ensuring coherence.

KB′=KB∪ΔK\mathbf{K}_B' = \mathbf{K}_B \cup \Delta \mathbf{K}KB′​=KB​∪ΔK

• Assessment: Evaluate discrepancies in ethical and contextual judgments between WA\mathbf{W}_AWA​ and WB\mathbf{W}_BWB​.

• Mathematical Measure: GapW=∥WA−WB∥\text{Gap}_W = \|\mathbf{W}_A - \mathbf{W}_B\|GapW​=∥WA​−WB​∥

• Transfer Method: Discuss ethical considerations and contextual decision-making strategies.

• Integration Mechanism: Infuse new wisdom into WB\mathbf{W}_BWB​, aligning with ethical frameworks.

WB′=WB+ΔW\mathbf{W}_B' = \mathbf{W}_B + \Delta \mathbf{W}WB′​=WB​+ΔW

• Assessment: Determine differences in Goals and Intentions between PA\mathbf{P}_APA​ and PB\mathbf{P}_BPB​.

• Mathematical Measure: GapP=∥PA−PB∥\text{Gap}_P = \|\mathbf{P}_A - \mathbf{P}_B\|GapP​=∥PA​−PB​∥

• Transfer Method: Clearly articulate the intended goals and objectives.

• Integration Mechanism: Align or adjust PB\mathbf{P}_BPB​ to reflect a shared or complementary purpose.

PB′=PB+ΔP\mathbf{P}_B' = \mathbf{P}_B + \Delta \mathbf{P}PB′​=PB​+ΔP

Complexity in interactions arises from the intricate pathways connecting DIKWP components. We quantify complexity ($\mathcal{X}$) as:

XAB=∑XγX⋅complexity(XAB)\mathcal{X}_{AB} = \sum_{X} \gamma_X \cdot \text{complexity}(\mathbf{X}_{AB})XAB​=X∑​γX​⋅complexity(XAB​)

Where γX\gamma_XγX​ are weights assigned to each component's complexity.

1. Simplified Communication Protocols:

• Standardization: Use standardized terminologies and structures to facilitate easier integration.

• Modular Messaging: Break down messages into smaller, manageable DIKWP components.

2. Hierarchical Integration:

• Top-Down Approach: Start with high-level Purpose alignment before delving into detailed components.

• Bottom-Up Refinement: Begin with Data and Information, progressively building Knowledge and Wisdom.

3. Iterative Refinement:

• Feedback Loops: Continuously refine DIKWP components based on feedback to streamline interactions.

• Adaptive Thresholds: Adjust thresholds (θX\theta_XθX​) dynamically based on interaction history and outcomes.

• Inconsistency Metric:

  IAB=∑XδX⋅inconsistency(XAB)\mathcal{I}_{AB} = \sum_{X} \delta_X \cdot \text{inconsistency}(\mathbf{X}_{AB})IAB​=X∑​δX​⋅inconsistency(XAB​)

  Where δX\delta_XδX​ are weights for each component's inconsistency.

• Resolution Mechanisms:

• Conflict Identification: Detect conflicting information or Knowledge.

• Reconciliation Processes: Implement methods to reconcile differences, such as seeking additional Data or Information.

• Uncertainty Metric:

  UAB=∑XϵX⋅uncertainty(XAB)\mathcal{U}_{AB} = \sum_{X} \epsilon_X \cdot \text{uncertainty}(\mathbf{X}_{AB})UAB​=X∑​ϵX​⋅uncertainty(XAB​)

  Where ϵX\epsilon_XϵX​ are weights for each component's uncertainty.

• Mitigation Strategies:

• Data Verification: Cross-validate Data components to reduce uncertainty.

• Information Clarification: Provide additional context or details to enhance clarity.

• Knowledge Updating: Incorporate new findings or evidence to strengthen Knowledge bases.

Maximize Understanding (UAB\mathcal{U}_{AB}UAB​) while minimizing inconsistency (IAB\mathcal{I}_{AB}IAB​) and complexity (XAB\mathcal{X}_{AB}XAB​):

max⁡UAB−λ⋅(IAB+XAB)\max \quad \mathcal{U}_{AB} - \lambda \cdot (\mathcal{I}_{AB} + \mathcal{X}_{AB})maxUAB​−λ⋅(IAB​+XAB​)

Where $\lambda$ is a penalty coefficient balancing the trade-off between maximizing Understanding and minimizing Inconsistency and Complexity.

To ensure effective communication, the following constraints are imposed:

CAB≥θC\mathcal{C}_{AB} \geq \theta_CCAB​≥θC​IAB≤θI\mathcal{I}_{AB} \leq \theta_IIAB​≤θI​XAB≤θX\mathcal{X}_{AB} \leq \theta_XXAB​≤θX​

Where:

• θC\theta_CθC​: Minimum required connectivity.

• θI\theta_IθI​: Maximum allowable inconsistency.

• θX\theta_XθX​: Maximum allowable complexity.

Employing optimization techniques such as Lagrangian multipliers or gradient descent to adjust DIKWP components:

1. Initialize: Start with initial DIKWP profiles for both stakeholders.

2. Iterate: Adjust DIKWP components based on gradients to maximize the objective function.

3. Converge: Continue iterations until constraints are satisfied and the objective function is optimized.

A medical practitioner (Human) collaborates with an AI diagnostic system (AI) to diagnose a patient's condition. Their interaction involves sharing and integrating DIKWP components to achieve an accurate diagnosis.

• Human (ProfileH\text{Profile}_HProfileH​):

• Data (DH\mathbf{D}_HDH​): Patient symptoms, medical history.

• Information (IH\mathbf{I}_HIH​): Identified potential diagnoses.

• Knowledge (KH\mathbf{K}_HKH​): Medical expertise and clinical guidelines.

• Wisdom (WH\mathbf{W}_HWH​): Ethical considerations in patient care.

• Purpose (PH\mathbf{P}_HPH​): Accurate and timely diagnosis.

• AI (ProfileA\text{Profile}_AProfileA​):

• Data (DA\mathbf{D}_ADA​): Extensive medical datasets, research articles.

• Information (IA\mathbf{I}_AIA​): Pattern recognition in patient data.

• Knowledge (KA\mathbf{K}_AKA​): Compiled medical knowledge from training data.

• Wisdom (WA\mathbf{W}_AWA​): Predefined ethical guidelines.

• Purpose (PA\mathbf{P}_APA​): Assist in diagnosis and treatment planning.

1. Data Exchange:

• Human to AI: Sends patient data (DH\mathbf{D}_HDH​) to AI.

• Gap Identification: AI identifies any missing Data components relevant for diagnosis.

2. Information Sharing:

• AI Analysis: Processes DH\mathbf{D}_HDH​ to generate Information (IA\mathbf{I}_AIA​).

• Shared Information: AI communicates patterns and potential diagnoses back to the Human.

3. Knowledge Integration:

• Human Review: Evaluates AI's Information (IA\mathbf{I}_AIA​) against personal Knowledge (KH\mathbf{K}_HKH​).

• Knowledge Update: Human incorporates relevant Knowledge from AI into KH\mathbf{K}_HKH​.

4. Wisdom Application:

• Ethical Judgment: Both stakeholders apply Wisdom (WH\mathbf{W}_HWH​ and WA\mathbf{W}_AWA​) to ensure ethical considerations in diagnosis.

5. Purpose Alignment:

• Goal Confirmation: Both entities confirm their Purpose (PH\mathbf{P}_HPH​ and PA\mathbf{P}_APA​) to ensure alignment towards accurate diagnosis.

• Mathematical Optimization: Adjust DIKWP components to minimize GapX\text{Gap}_XGapX​ and maximize UHA\mathcal{U}_{HA}UHA​.

• Result: Enhanced Understanding leads to a more accurate and reliable diagnosis, reducing the likelihood of AI-generated hallucinations (erroneous diagnoses).

The Theory of Relativity of Hallucination offers a mathematically grounded framework for understanding and enhancing communication between stakeholders through their personalized DIKWP*DIKWP profiles. By systematically identifying and sharing missing DIKWP components, and addressing connectivity, inconsistency, uncertainty, and complexity, stakeholders can achieve a higher degree of mutual Understanding. This framework not only mitigates the incidence of hallucinations in AI systems but also fosters effective human-AI collaboration, ensuring that interactions are coherent, reliable, and ethically aligned. Future research should focus on empirical validation of the proposed mechanisms and the development of advanced algorithms to facilitate seamless DIKWP*DIKWP interactions.

1. Duan, Y. (2023). Lecture at the First World Conference of Artificial Consciousness.

2. Duan, Y. (Year). International Test and Evaluation Standards for Artificial Intelligence Based on Networked Data-Information-Knowledge-Wisdom-Purpose (DIKWP) Model.

3. Vaswani, A., et al. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 5998–6008.

4. Sutton, R. S., & Barto, A. G. (2018). Reinforcement Learning: An Introduction. MIT Press.

5. Bengio, Y., LeCun, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444.

6. Silver, D., et al. (2016). Mastering the game of Go with deep neural networks and tree search. Nature, 529(7587), 484–489.

7. Marcus, G., & Davis, E. (2020). GPT-3, Bloviator: OpenAI's language generator has no idea what it's talking about. MIT Technology Review.

8. Maynez, J., Narayan, S., Bohnet, B., & McDonald, R. (2020). On faithfulness and factuality in abstractive summarization. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 1906–1919.

9. Kant, I. (1781). Critique of Pure Reason.

10. Seth, A. K. (2014). A predictive processing theory of sensorimotor contingencies: Explaining the puzzle of perceptual presence and its absence in synesthesia. Cognitive Neuroscience, 5(2), 97–118.

The author extends gratitude to Prof. Yucong Duan for his pioneering work on the DIKWP model and the Theory of Relativity of Consciousness, which have significantly influenced the conceptual framework of this analysis. Appreciation is also given to colleagues in cognitive science and artificial intelligence for their invaluable feedback and insights.

Correspondence and requests for materials should be addressed to [Author's Name and Contact Information].

Keywords: DIKWP Model, Theory of Relativity of Hallucination, Human-AI Interaction, Cognitive Enclosure, Mathematical Framework, Data-Information-Knowledge-Wisdom-Purpose, Hallucination Mitigation, Understanding Enhancement