From almost a decade, it has been the basic requirement of every reader to have strong vocabulary, make comprehension strong and understand the context of text easily. The purpose of our project, Reading Assistant for Transcript with Enhanced Searching (RATES), is to enhance learning aptitudes, knowledge, understanding and literal level of user while reading a document. It has been realized in a desktop and mobile application.Using this app, a user takes a text or pictorial image of document. For pictorial image, relevant thumbnail of image on internet with labels, has been provided. On the other hand, textual image passes through optical character recognition (OCR). User selects a content (word, technical terms, sentences, paragraphs) and application provides vocabulary contents (dictionary, synonyms, antonyms etc.), text-to-speech, short summary, translation, web search results and keyword based searching features based on selection of text contents. App submits the request of user after each selection of content to external application program interface (API). Based on response from API, user gets the required result within the application. It is believed that RATES will assist people in getting relevant and specific information in smart, systematic and efficient way.
- Safi Ullah
- Nida Akbar
- Imran Khan
Data Analytics, Data Science, Machine Learning, Natural Language Processing
In this project, we aim to develop a recommender system for toxic comments detection that will avoid sending lethal comments to the other users who belongs to a different religion. Due to increased usage of recommender systems, they have become a necessity. Current state of art research deals with E-Commerce, Advertising and Social Networks related domains that help to enhance revenues, to facilitate product purchasing, to recommend a friend but we aim to extend this research in religion based domain which will help to detect toxics comments based on different religious aspects and discard them before reaching on the other end. Therefore, one of the main contributions of this project would be to develop a large corpus for Recommender System (RS). We used Kaggle's dataset and to applied various classification algorithms. For evaluation measures, Accuracy, Precision, Recall, F1 and Area Under the Curve are used.