Graduation Project CSC496

 

CSC 496 – Graduation Project-I  (2-0-0)
Course Description:			This course is the first part of a sequence of two courses (CSC496 and CSC497) that in entirety constitute the BSc graduation capstone project. In this project, the student is expected to develop software for a specific problem by applying previously learned concepts and methods during the course of the project. In this course the student is typically expected to study the problem, see what others have done, perform the analysis, determine the requirements and suggest/design a solution. The project will culminate in a formal public presentation, and written documentation. Oral and written progress reports are required. The project topic may be provided by the faculty, by the student(s) or by the industry. The topic is subject to the departmental approval.
Prerequisite:			Student must have finished at least 100 hours in the BSc program
Textbook:			No textbook
Course Objectives:			This course is a semester-long team project, where students apply a broad range of skills learned in earlier courses, and demonstrate their competence in technical material, communications, and project skills
Project Abstract: Our proposed project, MyInterests ( a bachelor graduation project for Computer Since Students in King Saud University) , is a native mobile application that works online as a recommendation system for Instagram users and will reduce the time they spend by filtering out uninteresting images . It works by keeping track of an Instagram account and notifying the user of important and relevant content. To do this, My interest program analyzes the images on a user account, identifies the interesting and preferable ones, and finally reports the results back to her\him.      In this project, we will apply Support Vector Machine (SVM), a machine learning method, to classify images and to assess whether they are likely to be preferred by the user or not. A user’s preferences can be derived from previously liked images on his or her account, based on the fact that preferred images usually share specific features that can be extracted from their content, colors, captions, and location. In this case, we use Cloud Vision API as a tool to analyze preferred images and extract the relevant features. For example, Cloud Vision API can analyze images and identify any text or logos contained in them; this information can then be used as a clue to advertising images. Also, it supports face, object, and emotion recognition to determine the image category (selfie, nature picture, etc.). Furthermore, it reveals the names of companies and products in the images to identify whether the user is interested in specific brands or not. These properties can help us to classify users’ preferred images.        In addition to notifying the user of the recommended images, the MyInterests application can suggest images to the user depending on the places he or she likes as an optional function. To achieve this, we used the Instagram API to deal with the user account, and we intend to make MyInterests available as a mobile application for a popular mobile operating system (Android).