Detection, characterization and analysis of unsafe content on YouTube - BC-363

Preferred Disciplines: Computer Science, Machine Learning, Deep Learning, Data Mining ; Master’s or higher
Project length:  6-12 months (up to 2 units)
Desired start date: Q1 – 2018 or ASAP
Location: Vancouver, BC
No. of Positions: 1
Preferences: UBC and SFU (the intern needs to come to the office at least 3 days a week). Language: English or Bilingual
Company: BroadbandTV (BBTV)


About Company:

BroadbandTV (BBTV) is a Vancouver-based tech and media company and the 3rdr largest video property worldwide behind Google and Facebook. BBTV connects content owners, content creators, audiences and advertisers by developing some of the most advanced and innovative technologies in the online video space and helps creators of all sizes improve their channel quality and reach new audiences.  

Project Description:

The aim of this project is to identify content on YouTube that is not advertiser friendly (e.g., videos with violence, extensive sexuality, racism, etc). In particular, we focus on detecting unsafe content for children.

The project will involve developing proper signal processing and machine learning solutions to analyze metadata and other signals from videos (possibly audio and video info) to achieve the above goal. 

Background and required skills

Research Objectives/Sub-Objectives:

The objective of the project is to create a safer video platform by utilizing signal processing, machine learning and deep learning technicqies to identify and flag questionable content (ex. violence, hate, racisim, content unsuitable for children).


  • Machine learning
  • Statistical and regression models
  • Deep learning
  • Signal processing

Expertise and Skills Needed:

  • Machine learning models: 
    • Linear regression, decision tree, neural network, Bayesian network
  • Statistical and regression models
  • Deep learning :
    • Platform:   Keras, Tensorflow, PyTorch, Theano
    • Method: ANN, CNN, RNN
  • Database related stills:  SQL
  • Programing language:  Python, R or Matlab 
  • Other desired skills:
    • Signal and Image processing: OpenCV
    • Natural language processing
    • Background in Computer Vision/Computer Science


For more info or to apply to this applied research position, please

  1. Check your eligibility and find more information about open projects.
  2. Complete this webform. You will be asked to upload your CV. Remember to indicate the title of the project(s) you are interested in and obtain your professor’s approval to proceed.
  3. Interested students need to get the approval from their supervisor and send their CV along with a link to their supervisor’s university webpage by applying through the webform or directly to Andrea Globa at agloba(a)