Extracting Decision Dependencies and Decision Logic from Text Using Deep Learning Techniques

Extracting Decision Model and Notation models from text using deep learning techniques

“…Automated decision models extraction approaches from structured data have been proposed such as in (Bazhenova & Weske, 2016b) for process models and in (Bazhenova et al, 2016a;De Smedt et al, 2019) for logs. Unstructured textual sources however contain more decision logic information compared to logs or process models and have been studied to extract decision dependencies (Goossens et al, 2021b)and decision logic (Arco et al, 2021) using natural language patterns. More sophisticated Natural Language Processing (NLP) deep learning techniques have been introduced such as Bidirectional Encoder Representations from Transformers (BERT) (Devlin et al, 2018) or Bidirectional LSTM with a Conditional Random Field (Bi-LSTM-CRF) (Huang et al, 2015) showing promising results that NLP could help to automate decision model extraction from text (Li et al, 2020;Lopez & Kalita, 2017;Otter et al, 2020;Torfi et al, 2020).…”