New Technology Alerts on Artificial Intelligence and Blockchain

Daily updated information

One of the main missions of the Spanish Patent and Trademark Office (SPTO) is the spread of the Technological Information gathered in the documents of patents, utility models, models and industrial drawings, industrial designs, trademarks and other distinctive sings that the Office stores. One of the most useful tools for this purpose is the Technology Alert, that provide global daily updated information related to a particular technical topic. Four new alerts have been recently added to the list, all of them pertaining to the Artificial Intelligence and Blockchain category, so the total number is none other than 29 Alerts.

The alert Blockchain or chain of blocks focuses on the practice of making and signing notes on any kind of transaction and recording them in a distributed way using nodes. The transactions are verified with a mathematical Proof of Work (PoW) almost impossible to revert. This feature and its distributed nature make the blockchain work as a spread logbook, with thousands of identical copies and highly difficult to corrupt.

The alert Machine learning or automatic learning refers to a group of technologies and techniques that allow IT systems to learn in an automatic way from a series of given data. The learning can be supervised or non-supervised and its rhythm and accuracy can depend on whether new data are added to allow the system to improve its models. Most of Artificial Intelligence technologies use the machine learning.

The alert Neural networks focuses on an automatic learning IT system that reproduces the animal neural working when it comes to network processing information. In deep neural networks (deep learning) a part of the process is opaque to the human supervisor, so the way the system learns is unknown. This group of techniques is used preferably for classification or pattern recognition problems, that are typical in the animal advanced intelligence.

The alert Fuzzy logic includes the techniques and implementations used to solve those classification problems that cannot be reduced to a total truth (for instance to decide if a colour is green or blue if chromatically they are very close). The classic logic is just able to state if something is True or False while the fuzzy logic can distinguish between veracity or falseness ranks.