Request for arXiv Endorsement for CARL Thesis (ConceptNet-Integrated AI Research)

[JoeSox]

Dear ConceptNet community,

I hope this message finds you well. I’ve been a longtime subscriber to this group and an early supporter of ConceptNet. Over the years, I’ve contributed articles and developed Microsoft C# integrations with ConceptNet for use in various AI projects — most notably my work on CARL (Cognitive Architecture for Reasoning and Learning), a biologically inspired humanoid AI framework that deeply integrates MBTI modeling, emotional simulation, and semantic reasoning using ConceptNet.

I’m now preparing to submit my formal thesis on CARL to arXiv.org under the eess.SY (Systems and Control) category and noticed that some of you are eligible to endorse submissions in this field.
Would you be willing to review and consider endorsing my submission, or kindly point me toward someone who can?

📄 Resources for Review:

Draft PDF: CARL Thesis – Dropbox Link https://www.dropbox.com/scl/fi/1hjm9s6nsl1q0foiwqp6h/CARL-Thesis_arxiv_working.pdf?rlkey=7f7mbuurp6m6cktam5uybuffd&st=lip9a2z4&dl=0 

arXiv Endorsement Link: https://arxiv.org/auth/endorse?x=FL6YAS


This work builds upon several open-source AI tools and semantic knowledge bases like ConceptNet and seeks to contribute a measurable architecture for conscious software, with strong alignment to the goals of transparent, explainable, and socially capable AI.
I'm sharing ongoing updates and resources here as well:
🌐 https://carl.earthbotics.com 


I greatly appreciate your time and consideration. I fondly remember supporting your IT and research infrastructure during my time at PNRI — it was a great experience working with such a dedicated group of scientists and engineers.
Thank you in advance, and please don’t hesitate to reach out with any questions or feedback.

Snippet from the embodied AI thesis related to ConceptNet:

"Commonsense reasoning combines Gordon & Hobbs' Accessibility-by-Association typed axioms with ConceptNet.
2.4.4 Commonsense Reasoning (v5.15.0)
The commonsense reasoning system introduced in version 5.15.0 implements Gordon & Hobbs' "Accessibility by Association" framework (2004), providing strategic planning commonsense through typed axioms that encode general knowledge about actions, events, and their relationships. This framework utilizes preconditions, effects, sequences, and abnormality markers to represent structured knowledge about how actions interact with the world, enabling CARL to reason about consequences and plan sequences of actions. AccessibleSet retrieval employs efficient concept accessibility scoring and ranking algorithms that identify relevant knowledge based on semantic proximity and contextual relevance, allowing rapid access to pertinent commonsense information. Plan support building leverages these accessible concepts to construct strategic reasoning sequences that reflect commonsense principles about goal achievement and action sequences. ConceptNet integration expands the knowledge base through external knowledge sources, providing validation and enrichment of internally maintained commonsense knowledge. This integration enables CARL to verify assertions against established semantic relationships in ConceptNet's extensive knowledge graph, ensuring that reasoning aligns with general world knowledge. The combination of typed axioms and ConceptNet validation creates a dual-source commonsense reasoning system that balances internal structured knowledge with external semantic validation."

--

Thank You,

Joe Socoloski III

DBA: Earthbotics