Abstract: In some embodiments, the techniques described herein relate to systems and methods to identify merchants from noisy transaction data with low latency and cost. The method receives transaction text and a zipcode, applies rule-based matching, and upon failure uses enhanced string distance (ESD). If ESD fails, a transformer model is invoked: an encoder-only model that embeds transactions and merchant names for similarity ranking, or a decoder-based model that generates a canonical merchant name with a confidence score. Candidates are filtered using zipcode and text search, and a merchant is selected by combining the generated name or embeddings with similarity metrics. Verification integrates name similarity and model confidence, and, when thresholds are met, enhanced merchant information is stored in an enhanced transaction database. The approach improves accuracy and coverage for real-time transaction understanding while reducing rule maintenance and large-model burdens.

Abstract: System and method for generating time series forecasts based on probabilistic data are disclosed. A processor receives a request to identify at least one feature of a dataset that is most closely correlated to a single feature specified in the received request. The dataset includes the probabilistic data of the time series. The processor identifies a set of original features within the dataset and derives a degree of dependency between a single specified feature and each original feature by using a temporally first portion of the dataset. After deriving a degree of dependency between all the features including the single specified feature and each engineered feature of the set of original features, the processor identifies an original feature or an engineered feature with the highest degree of dependency as the at least one feature of the dataset that is most closely dependent to the single specified feature.