Trinity College Dublin

<p><strong>Overview:</strong> As Ireland’s university on the world stage, Trinity is recognised for academic excellence and a transformative student experience. The historic campus is located in the heart of Dublin city centre at the meeting place of the retail and cultural districts. With a tradition of scholarship spanning more than four centuries, Trinity is home to talented and inquiring minds, a liberal education, and research conducted at the frontiers of disciplines.</p>
<p><strong>Research:</strong> Trinity’s position as one of the world’s leading universities is defined by our research, scholarship, and capacity to address issues of global, societal and economic importance. Our strength lies in the breadth and excellence of our expertise; from individual-scholars and thought-leaders, to large multidisciplinary teams, each delivering research of national and global importance.</p>
<p>Trinity’s research themes address issues of immediate and long-term concern for society and offer opportunities for future economic development. They are supported by a set of research institutes that provide the infrastructure that is required to support multi-disciplinary research as well as partner engagement.</p>
<p><strong>Industry Partnerships:</strong> Trinity is committed to supporting industry — from SMEs to MNCs, from short term projects to long term on-going collaborative projects. In a national or international context, we work collaboratively with industry stakeholders and agencies to share infrastructure and expertise, tackle industry relevant research challenges, problems, or address grand challenges.</p>


<h1 id="background">Background</h1>
<p>In 2017, an airline was launching a new accommodation booking service. They had limited data on guests using this new service and wanted to explore how it could offer a more personalised experience to these customers. The two main approaches to recommender systems are Collaborative Filtering and Content-based approaches, which do not perform well where there is limited data.</p>
<h1 id="technology-overview">Technology Overview</h1>
<p>A real-time hybrid recommender that combines different techniques and exploits all the available information about users, such as: </p>
<ul>
<li>User preferences to personalise recommendations</li>
<li>Group preferences to capture preferences of similarly grouped people</li>
<li>Data associated with items to apply content-based techniques</li>
<li>Contextual Information</li>
</ul>
<p><img src="https://s3-eu-west-1.amazonaws.com/assets.in-part.com/technologies/figures/uk8ewb9tSQimitXwL7Rj_Picture13.png" alt="Figure 1" /></p>
<p><strong>Inputs:</strong> User preferences and contextual information.</p>
<p><strong>Adapt Technology</strong>: Hybrid approach that blends elements of naïve collaborative filtering, content-based recommendation and contextual suggestion.</p>
<p><strong>Outputs:</strong> Custom accommodation recommendations for guests using the service.</p>
<h1 id="benefits">Benefits</h1>
<ul>
<li>A recommender that overcomes the cold start problems of content-based and collaborative filtering approaches</li>
<li>Deployed with the airline for evaluation</li>
<li>Does not require significant rating data</li>
<li>Generates personalised recommendations</li>
<li>Provides real-time and robust recommendations</li>
</ul>
<h1 id="opportunity">Opportunity</h1>
<ul>
<li>License</li>
<li>Research collaboration</li>
</ul>
