Skip to content
SolutionTogether Ecosystem 1.1
Visualize Topics
Brainstorming
Projects
Project Management
Funding 1.1
Bevisioneers
Thesis 1.3
Adam's project (John’s extraction notes)
Atlantis Together - Research & Management
Untitled sync page
Share
Explore

SolVec

Executive Summary
SolVec is an innovative solution/proposal assessment API that utilizes cutting-edge natural language processing (NLP) and knowledge graph technologies to provide comprehensive insights into the feasibility and potential impact of various initiatives. By leveraging a vast repository of open research papers, SolVec identifies and analyzes patterns in data to assess the effectiveness of proposed solutions and inform decision-making processes. This groundbreaking tool aligns with the EU's strategic goals of promoting interoperability, adopting decentralized technologies, and harnessing AI for societal progress.
Project Description
SolVec's core functionality revolves around extracting, analyzing, and interpreting data from a comprehensive dataset of research papers. The initial step involves filtering the dataset to focus on papers that demonstrate a clear impact on specific indicators, quantified using betweenness centrality measures. This filtering process ensures that SolVec only considers data that is relevant to the assessment of proposed solutions.
Next, SolVec utilizes NLP techniques to extract relevant attributes from the selected research papers. These attributes include topic labels, funding information, nouns, indicators, their dimensions, growth ranges and certainty levels, reasons for the observed effects, experimental methods, and sample sizes. This comprehensive data extraction ensures that SolVec has a deep understanding of the underpinnings of various solutions.
The extracted data is then vectorized and uploaded to a decentralized, unstoppable vectorial database, ensuring data integrity and accessibility. This decentralized approach aligns with the EU's focus on interoperability and the adoption of decentralized technologies.
To further enhance the capabilities of SolVec, a knowledge graph is generated using LLaMA, an open-source AI language model. This knowledge graph provides a structured representation of the relationships between entities, connecting the extracted attributes and enabling deeper insights into the underlying data patterns.
A RAG script is developed for LLaMA, enabling the evaluation of proposed solutions against the extracted patterns. This RAG script assesses the proposed solutions based on their alignment with proven tendencies and their likelihood of supporting the set tendencies within the context of the extracted patterns.
A web-based user-friendly interface (UI) is created to allow users to specify the subject, indicator, and desired effects they are interested in, and SolVec provides insights into the proven tendencies and potential impact of various proposals.
An open-source ethical API is developed to predict the potential consequences of decision proposals, including potential impacts on social and environmental sustainability. This API complements SolVec's effectiveness in evaluating the feasibility and impact of proposed solutions by default, quickly on a large scale.
To ensure a seamless user experience, the UI and API features are integrated, allowing users to seamlessly navigate between the two platforms and obtain a comprehensive assessment of proposed solutions.
Impact and Significance
SolVec holds immense potential to revolutionize the way solutions are assessed and decision-making processes are informed. By providing an unbiased and comprehensive assessment of the feasibility, potential impact, and ethical considerations of proposed solutions, SolVec can significantly improve the quality of decision-making across various sectors, including social, environmental, and economic domains.
Results can be used to fact-check scientific results in Wikipedia articles.
⁠
image.png
⁠
⁠
⁠
NGI application by SolVec.pdf
161.2 kB
⁠
⁠
Load content from docs.trace.market?
Loading external content may reveal information to 3rd parties. Learn more
Allow
Load content from package.trace.market?
Loading external content may reveal information to 3rd parties. Learn more
Allow
Want to print your doc?
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.