IASTEM INTERNATIONAL CONFERENCE
Paris, France 25-04-2022
2 papers
-
Do Words in U.S. 10-K Filings Reflect Diversity and Inclusion Efforts?
Benjamin P Foster, Andrew Manikas
Paper ID IA-ICEBMPARI-250422-15093 · Pages 24-26
Download PDF: Do Words in U.S. 10-K Filings Reflect Diversity and Inclusion Efforts?Abstract
Studies have examined the impact of diversity on firm operations and performance. Diversity has become a value encouraged by society and a major strategy and goal for many companies. This study examines the number of unique diversity words mentioned in Management’s Discussion and Analysis, Item 7 of 10-Ks filed with the Securities and Exchange Commission of the U.S. We examine if the unique number of diversity related words mentioned in the 10-Ks can provide a measure of a company’s commitment to diversity by comparing the diversity word count to measures of diversity provided by independent outside organizations. The unique diversity word counts from Item 7 are significantly correlated with diversity measures/ratings constructed by outside entities. Keywords - 10-K filings, Corporate Equality Index, Diversity and Inclusion.
-
Application of Computer Vision and Artificial Intelligence in Automated Document Processing
Zrinka Gligo, Sven Gotovac
Paper ID IA-ICCARVEPARI-250422-15122 · Pages 27-30
Download PDF: Application of Computer Vision and Artificial Intelligence in Automated Document ProcessingAbstract
With the increase of computer power and a major expansion of available data, computer vision is rapidly developing as well as its application in everyday needs such as advertising, financial statements, surveillance, medicine, robotics, self-driving vehicles, etc. Handwritten digit recognition is a popular problem and numerous papers achieve excellent results in solving it. The goal of this work is a similar problem, the recognition of stamped digits in a form of a date. This paper proposes an approach that includes frequently used algorithms in handwritten digit recognition like neural network as well as specific image pre-processing suitable for our particular data. Due to high level of accuracy achieved,  the solution found application in the automatization of the billprocessing. Keywords – Convolutional Neural Network (CNN), Digit Recognition, Image Classification, Keras, OpenCV