Feature Engineering Over Unstructured Data
This research aims to advance the application of traditional artificial intelligence (TAI) to unstructured data. The key contributions include the development of a novel, effective, and efficient pipeline that dynamically constructs a traditional feature space amenable to TAI algorithms. The constructed feature space exhibits accuracy and precision across numerous disparate TAI algorithms, ensuring wide applicability. Furthermore, the study introduces an innovative implementation of the extended De Bruijn graph, enhancing the efficiency of path discovery. The presented approach is not only robust but also scalable and modular, which enhances its potential for diverse applications and future adaptations.