Automatic Extraction of Communicative Signals from Online Flowchart

Abstract

Flowchart infographic used as an effective learning tool to understand the steps involved in a process and conveys message through set of symbols with text arranged in specific structure. The need of accessible flowchart remains critical in case of its usage by visually impaired. The objective of paper is to automatically extract and identify the types of communicative signals appear in a flowchart to understand the structure. The paper also includes brief about how to exploit these communicative signals to infer the intended message for visually impaired. The use of various communicative signals including symbolic metadata, explicit textual message, and perceptual task to construct the knowledge is demonstrated through examples. The resulting intended message gives the structural knowledge of flowchart as well as the simplified description of flowchart.  Performance evaluation is done for accuracy of automatic extraction of communication signals present in online flowchart with the accuracy of 79% for shapes and 67% for flowline.