21 October 2020 in International Journal of Automation and Computing
28 May 2021 in IEEE Access
World Wide Web has redefined almost all the business models in the past twenty-five to thirty years. IoT, Big Data, AI are some of the comparatively recent technologies which brought in a revolution in the digitization and management of data. Along with the revolution arose the need for data security and consumer privacy protection, primarily concerning financial institutions. The data breach of Equifax in 2017 and personal information leaks from Facebook in 2021 led to general skepticism among the customers of large corporations. The GLBA, 1999, also known as the Financial Modernization Act, was implemented by US federal law to enforce the financial institutions to protect their private information. Built upon the GLBA, guidelines are paved by FTC for all financial institutions of the United States of America, including TI companies. In this paper, an ANN-based content classification technique using MLP architecture in combination with n-gram TF-IDF feature descriptor is proposed to detect and protect the customers’ sensitive information of a reputed TI company securing its one of the digital image-document stores. The proposed technique is compared with other state-of-the-art strategies. Data samples from the digital document store of the company have been taken into consideration in the study, and the prediction accuracy metrics obtained are found to be substantially better and within the acceptable range defined by the organization’s information security monitoring team.
20 January 2022 in IEEE Access
In the twenty-first century, storing and managing digital documents has become commonplace for all corporate and public sectors around the world. Physical documents are scanned in batches and stored in a digital archive as a heterogeneous document stream, referred to as a digital package. To make Robotic Process Automation (RPA) easier, it’s necessary to automatically segment the document stream into a subset of independent, coherent multi-page documents by detecting the appropriate document boundary. It’s a common requirement of a TI company’s Automated Document Management Systems (ADMS), where business operations are automated using RPA and the goal is to extract information from digital documents with minimal user intervention. The current study proposes, evaluates, and compares a multi-modal binary classification network incorporating text and picture aspects of digital document pages to state-of-the-art baseline methodologies. Image and textual features are extracted simultaneously from the input document image by passing them through Visual Geometry Group 16 -Convolutional Neural Network (VGG16-CNN) and pre-trained Bidirectional Encoder Representations from Transformers (Legal-BERTbase) model through transfer learning respectively. Both features are finally fused and passed through a fully connected layer of Multi Layered Perceptron (MLP) to obtain the binary classification of the pages as the First Page (FP) and Other Page (OP). Real-time document image streams from production business process archive were obtained from a reputed Title Insurance (TI) company for the study. The obtained F1 score of 97.3797.15the expected Straight Through Pass (STP) threshold defined by the process admin.