The MATRIX Semantic Analysis Engine: Introducing Intelligent Contracts

MATRIX auto-coding Intelligent Contracts outpace other smart contracts by supporting easy-to-use, customizable, high-level functions. This is made possible by the MATRIX Semantic Analysis Engine.

To put it bluntly, the MATRIX AI Network is building a blockchain that people will actually be able to use. This requires faster, safer and more intelligent technology. But cutting-edge technology, on its own, isn’t enough. A healthy, responsive and lively ecosystem is also necessary for a platform to grow strong.

MATRIX’s goal is to create a platform that enables Artificial Intelligence and blockchain technology to work in unison in order to improve the lives of users. Frankly, this is a lofty goal. To achieve something of this scope and nature, the importance of prioritization is absolute. As such, MATRIX is targeting several glaring shortcomings present in current blockchain solutions.

Front and center are the programming barriers erected by the unnecessarily convoluted first-generation smart contracts introduced by the Ethereum network. While the entire blockchain industry should be thankful to Ethereum for popularizing the concept, Ethereum smart contracts are, quite simply, not smart enough.

The MATRIX AI Network Solution

To address the unnecessary and damaging programming barriers plaguing the development of DApps and, ultimately, the adoption of blockchain solutions, the MATRIX AI Network is designing proprietary Intelligent Contracts that leverage natural language programming and adaptive deep learning-based templates to auto-code. Currently, English and Chinese are supported. In the future, support for other languages will be added according to popular demand.

MATRIX’s auto-coding Intelligent Contracts have two major innovations. The first is the MATRIX Semantic Analysis Engine. The MATRIX Semantic Analysis Engine handles the automatic generation of Intelligent Contract code and is the topic of this write-up. The second is the MATRIX Secure Virtual Machine. The MATRIX Secure Virtual Machine is used to detect security vulnerabilities and will be introduced in greater detail in a subsequent article.

The MATRIX Semantic Analysis Engine: Automatic Generation of Intelligent Contract Code

Existing smart contracts, despite their name, are not particularly smart. Nor are they particularly convenient. They are essentially just collections of static code that run on blockchain networks. MATRIX’s auto-coding Intelligent Contracts, on the other hand, are incredibly easy to use. The user simply types or speaks the features that they want to include. Then, the Intelligent Contract automatically handles the coding.

MATRIX’s adaptive deep learning-based templates differ from other early attempts at smart contract templates in their complexity. Other smart contract templates that supported one-click contract generation typically only let users define very basic information such as a token’s name or a token’s total supply. Other high-level functions are typically not supported. This has directly hindered the adoption of smart contracts. MATRIX’s Intelligent Contract prototype currently supports One-Click Token Launch, One-Click ICO, Voting, Insurance and Auctions; just to name a few.

The MATRIX AI Network team has achieved this feat by conducting extensive semantic analysis of countless open source smart contracts to develop a proprietary deep learning model. This proprietary deep learning model allows the MATRIX Semantic Analysis Engine to perform semantic analysis on Intelligent Contracts. First, an abstract semantic analysis is conducted to generate a control-flow graph. Then, semantic features are extracted from the control-flow graph and are used to continuously train MATRIX’s deep learning model to identify specific patterns and parameters, as well as to identify and define clusters of semantic features. By doing so, the MATRIX AI Network’s Intelligent Contracts are able support multiple high-level functions.

The MATRIX Semantic Analysis Engine prototype already supports customized, high-level token functions relating to transfers, airdrops, mining, burning and more. Other functions relating to ownership and management rights, account freeze-and-suspend functions and more are also currently supported.

While more technical details about how this all works will be shared at a later date, the MATRIX AI Network team has prepared a short demonstration for the MATRIX community. Enjoy!


During the month of September, MATRIX AI Network is releasing a series of videos and write-ups highlighting some of the project’s technical features. These include the MATRIX Mining Machine, the MATRIX Auto-Coding Intelligent Contracts, the MATRIX Secure Virtual Machine and the MATRIX Digital Asset Safe.