AI and BlockChain Technology Research and Development at SoftSim Technologies Inc.
SoftSim Technologies Inc. has had a focus on the development of technologies that improve scientific research, more specifically medical. In recent years, research has shown that the scientific research industry is increasingly moving towards the utilizing of latest advancements in Blockchain, Deep Learning, AR and VR using the powerful GPU processing power.
As such, SoftSim Technologies conducted its own research, focusing on the extraction of data from the accessible Big Data reserves (available on the World Wide Web)using deep learning and AI techniques. Extracted data is then classified and synthesized to researchers. In addition, the research also focused on combining deep learning algorithms and blockchain smart contracts to manage confidential data. Due to the fact that blockchain infrastructure and technology was designed for bitcoin and crypto-currency, it needs to be adapted for management of smart contracts that are not based necessarily on financial compensations.
During this research, some of the challenges faced included: measuring the accuracy of the data sources. In addition, some uncertainties faced included (listed below a few):
- Which cryptographic algorithms to use to secure data ?
- What are the most efficient methods to break down and classify the medical data in different blockchains based on type?
- Which techniques and methods to use to implement the smart contracts
Uncertainties were managed through continued building and testing of new prototypes ( in order to enhance the performance and achieve the required level of data security), technical experimentations (such as testing of OpenSemantic Platform, testing methods and encryption algorithms) and development/ testing of prototyped smart contracts (to be used for a typical clinical trial life cycle).
SoftSim advanced in this technical evaluation considering several key aspects, including but not limited to: the continuous improvement of the performance and capabilities of this prototype within our core technology set. In addition, we conducted analysis, testing and problem resolutions.
Our experiment led us to especially focus on the advancements Blockchain supporting Electronic Medical Records would bring: agility and efficiency, security, simplicity, scalability, improved data quality.
AppHolo FAQ Design Document
It will start with an overview of the AppHolo project and the control schemes for the FAQ application. After which the user will be in a room where they can click on various items to see and hear a more in depth explanation of that piece of the AppHolo project. This explanation can bring the user outside the room.
- Use an unspecified button (maybe Y button) to teleport to locations in the room.
- To ensure that the user can skip the current explanation, the user can press a currently undefined button (maybe B button) to skip the explanation and go back to the room.
First a short presentation overview of AppHolo is played, which the user can skip and at the end give an explanation of the core mechanics, and what questions can be answered. The user is then placed into a room with various objects representing various aspects of the AppHolo project. The user can interact with these objects to gain further understanding of the AppHolo project.
Data for Axcess 1
SoftSim Technologies needs training data to train an AI for identifying what information is public knowledge in company and government documents. SoftSim Technologies is building an AI system to help users identify information in documents that are potentially classified and what information is public knowledge. To help identify information that is public knowledge SoftSim is planning on using an AI, however this AI first needs to be trained on such a task. This document’s aim is to secure such training data. In particular, SoftSim wishes to record the discovery process of finding information that is public knowledge, particularly what document data is linked to what found data.
Areas of Use
The AI can have additional benefits and uses outside of finding public information in documents. The AI essentially finds information links between paragraphs and documents. This information linkage can have various uses. These information links can help create a graph between various documents within an organization. Documents with a similar topic (possibly used for different audiences and purposes) will have high information linkage. This can then be used to meaningfully group documents together on a topic.
This linking of information across documents can also help develop strategic planning of future projects. Information linkage in documents across different departments can help identify common assets and interests across these departments, and thus affect strategic planning of projects involving these assets and interests. If a project builds upon assets covered in several different documents, information linkages between these documents can give a clearer understanding of these assets. A greater understanding of these assets can also help reduce duplicate tasks for the project.
These information linkages can also help a user find additional information on a topic. For example, if there is an information link between two documents and the user is reading one document, the information linkage between the documents can lead the user to new information within the second document.
The basic control scheme is to:
- Click on an object to show animation explanations associated with clicked object (index finger button)
- Alternatively, a user can verbally ask questions about the projects: The start of a question can be created by a wake word or an unspecified button press (maybe X button.
- Hover the cursor over an object to see the text description of the object.