Year over year, the number of books read by both children and adults has been on the decline while the average number of minutes spent on mobile devices has skyrocketed. This is evidently reflected in retail bookstores transitioning from traditional paper-based books to encompass more digital products including e-readers and tablets. Just as Gutenberg’s printing press has come and gone, physical books, newspapers, and textbooks face an inevitable future. In an era where everything is expedited and online, the act of reading consumes too much time to embrace as a pastime and is often lost in the pace of society.
The disconnect between reading lengthy books and the digital age of immediate information consumption is reconciled with an intermediary: Summary Scanner. Hoping to counter the movement that shifts away from reading, the Summary Scanner app allows users to get an automatic summary of any document in seconds. With the phone’s camera, Summary Scanner uses OCR and Natural Language Processing (NLP) to determine which sentences are most relevant in providing an overview of the document without dismissing important information.
Instead of the disenchanting process of going through a stack of papers to gather key points, an image of the text is all that is needed to generate an easy-to-read summary. Beyond summarizing, the app also allows users to convert physical documents to form editable, searchable electronic versions to translate, use text-to-speech, or share. The inception of Summary Scanner began when Kevin, 21, purchased a $200 university textbook only to open it a total of 2 times during the entire semester as a result of the time it would take to go through it. This situation is not unusual for thousands of students, where textbooks remain untouched as reading becomes too time consuming. Realizing that a solution had to be created for the countless of students facing a similar dilemma, he began exploring hyper-efficient methods of information use.
With the Android version currently available on the Google Play Store and the iOS version in production, Summary Scanner is hoping to disrupt how information is consumed in schools and professional environments. Given that access to mobile devices is increasingly used for productivity and work, its use is estimated to increase drastically during the school year as students find more efficient ways to get ahead of their peers. Being first to market with such a product, Summary Scanner is poised to take the lead against other educational tools aimed at increasing productivity.
Summarization Algorithm – How Summary Scanner Works
As the core mission of the app is to increasing reading efficiency, Summary Scanner’s automatic summarization algorithm utilizes Natural Language Processing (NLP), a rapidly growing field within computer science, to derive meaningful summaries from raw text. Through machine learning, the app first associates text with similar definitions (e.g. “house”, “houses”, “home”) to determine the frequency usage of a particular term. The detection and processing of terminal punctuation marks (periods, question marks, exclamation marks, etc.) which are generally used to end sentences must be then correctly identified as such while distinguishing between non-terminal usages such as “Mrs.” and “U.S.”. Further analysis is done to identify main themes or topics, one way being to conduct keyword identification which distinguishes capital letters for names, places, dates, or the relative location of the text (title text in contrast to body text). With extraction-based summarization, Summary Scanner then returns the most significant concepts after removing transition phrases and reorganizing the summary to orient around the topic determined by the keywords.