According to GlobalData’s company profile on Worldline, contactless card payments was a key innovation area identified from patents. Worldline's grant share as of September 2023 was 52%. Grant share is based on the ratio of number of grants to total number of patents.

Machine learning system for text mining to detect defects or anomalies in computer applications

A recently granted patent (Publication Number: US11763137B2) describes a machine learning system designed to detect defects or anomalies in authentication, transactions, or operations carried out by computer applications. The system includes a hardware and software arrangement that forms a pre-processing system, which generates aggregated enriched data by quantifying the degree to which each authentication, operation, or transaction complies with previous ones in a temporally ordered sequence. This pre-processed data is then fed into a recurrent neural network of the long short-term memory (LSTM) type, either alone or in combination with a decision tree algorithm for statistical learning. The output of the recurrent neural network is used to validate the authenticity of the sequence of authentications, operations, or transactions.

The recurrent neural network in the system consists of at least two recurring layers and a Logistic Regression Classifier positioned above the last recurrent layer. The Logistic Regression Classifier takes into account the time elapsed between two authentications, operations, or transactions in the sequence. Additionally, the system includes a hardware and software arrangement that parameterizes the validation process using a Jaccard Index matrix, which measures the similarity between the output data of the LSTM neural network and the output data from a second neural network for statistical learning. This allows for the validation of results from either of the two neural networks.

The system is specifically configured to operate as a computer application for risk prediction and fraud detection in authentication operations of electronic memory objects containing secret information. It utilizes a recurrent neural network of the LSTM type, which may include a GPU for enhanced processing power. The pre-processing system of the machine learning system includes at least one first database containing sequential patterns of raw data, a second database containing external data, and arrangements for enriching and aggregating the data. The pre-processing system is also designed to utilize multi-threading for efficient processing.

Overall, this patented machine learning system offers a comprehensive approach to detecting defects or anomalies in computer applications' authentication, transactions, or operations. By utilizing a combination of pre-processing techniques, recurrent neural networks, and statistical learning algorithms, the system aims to improve the accuracy and efficiency of fraud detection and risk prediction in electronic memory objects.

To know more about GlobalData’s detailed insights on Worldline, buy the report here.

Data Insights

From

The gold standard of business intelligence.

Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.

Be better informed