StagePage Artist Similarity Network- ON-408

Discipline(s) souhaitée: Génie - informatique / électrique, Génie, Informatique, Sciences mathématiques
Entreprise: StagePage Ltd
Durée du projet: 4 to 6 months
Preferred start date: As soon as possible.
Langue exigée: English
Emplacement(s): Toronto, ON, Canada
Nombre de postes: 1
Établissements préférés: Université Ryerson, Université de Toronto, Université York

Rechercher dans les réseaux internationaux de Mitacs - cochez cette case si vous souhaitez recevoir des profils de chercheurs basés à l’extérieur du Canada: 
No

Au sujet de l’entreprise: 

StagePage is a performing arts intelligence company and data platform on a mission to connect audiences with great performance through personalized recommendation and matching solutions. We aim to: a) Increase the discoverability and access to the works of Canadian artists, both at home and abroad; b) Empower diverse publics to easily filter and discover work that matches their specific needs and values and c) Increase the capacity for research and analysis on audience behaviour and the changing landscape of the performing arts sector through national data collection and dissemination.

Veuillez décrire le projet.: 

We are looking to work with a data alchemist on a unique opportunity in the performing arts domain. The project will involve building a poster that demonstrates the similarities between performing artist creators in Canada & the US (including dance, opera and theatre creators). The goal will be to map creators' works based on a series of “Experience Personality” dimensions. Each creator’s personality is comprised of Genre, Sub-Genre, Style, Sentiment, Theme and Language. 

The goal of this project is to help StagePage build intelligence on the performing arts sector, so that we can improve recommendations for users and empower more people to experience and access the arts.

The main goal of the company is a recommender system called StagePage Recommends which helps users more easily find, discover and navigate potential arts experiences, and in turn, helps the content creators better share and promote work to the most suitable audiences.

This challenge will involve: 

  • Scraping source data from websites, media and performing arts databases 
  • Developing an information representation strategy (using knowledge representation & reasoning) in collaboration with the StagePage team to conceptualize dimensions of the Experience Personality. This step will involve answering the questions, what genres and sub genres exist in the performing arts? How do we know whether an arts experience fits into one of these categories? This step may include methods such as topic modelling.
  • Building a Natural Language Processing (NLP) script that can look for patterns in shared descriptive vocabulary between the creator’s works, and which can process text (e.g. a show review, program or directors notes), and categorize the experience personality dimension scores of the work automatically
  • Map similarities between performing arts creators, based on the experience personality scores of their associated work.
  • Create a data visualization that shows the similarities and differences of all creators. This step may take the form of an interactive app, or a poster including a data visualization graphic (Example http://labs.polsys.net/playground/spotify/).

Algorithmic techniques such as natural language processing and knowledge representation and reasoning.

 

Expertise ou compétences exigées: 

  • data processing, scraping and analysis
  • natural language processing
  • knowledge representation and reasoning