PORTALE DELLA DIDATTICA

Ricerca CERCA
  KEYWORD

UI/UX design and implementation of an e-mail client exploiting supervised learning techniques for messages’ organization

azienda Thesis in external company    


keywords EMAIL, MACHINE LEARNING, SUPERVISED LEARNING, USER EXPERIENCE, USER INTERFACE

Reference persons LUIGI DE RUSSIS

External reference persons Roberto Bressani (DNDG S.r.l., thesis@dndg.it)

Research Groups DAUIN - GR-10 - Intelligent and Interactive Systems - e-LITE

Thesis type EXPERIMENTAL

Description The two most used email clients in the world are Apple Mail and Gmail, according to the latest data (November 2022). Apple and Gmail together make up a whopping 88.57% of the total email client market share. The third-most popular email client is Microsoft's Outlook. It's used by 3.35% of email users worldwide.

The Gmail interface makes Gmail unique amongst webmail systems for several reasons. Most evident to users are its search-oriented features and means of managing email in a "conversation view" that is similar to an Internet forum. An official redesign of the Gmail interface was rolled out on November 1, 2011 that simplified the look and feel of Gmail into a more minimalist design to provide a more consistent look throughout Google products and services as part of an overall design change. Another major redesign took place in April 2018 which introduced new information rights management controls designed for business use cases.

Apple Mail (officially known as Mail) is an email client included by Apple Inc. with its operating systems macOS, iOS, iPadOS and watchOS. Apple Mail grew out of NeXTMail, which was originally developed by NeXT as part of its NeXTSTEP operating system, after Apple's acquisition of NeXT in 1997. The Apple Mail interface is based on a common Tree pattern, a visualization of a hierarchy. In the most typical orthogonal 2D layout of the Tree, the hierarchy level is shown by indentation on the left side of the Tree nodes. Emails can be grouped in folders or set and navigated through the tree.

Both email clients have not had major evolutions from the point of view of the user experience. On the one hand, Gmail puts the search function at the center of the user experience. However, searching mailboxes with thousands of emails is often complicated and it is difficult to find specific messages if you don't remember exactly the content of the conversation you are looking for.

On the other hand, Apple Mail uses more traditional UI/UX patterns allowing the user to organize emails into folders. However users tend to change their approach to organizing information over time and remembering the process by which the tree was organized becomes more complex as trees work best when users already know the hierarchy.

The aim of this project is to analyze the current state of the art of email clients and design a new client that exploits machine learning techniques to improve user experience and mail management over time. Our purpose is to design a client that allows you to best combine the search and folder organization function or novel UI/UX patterns using recommendation systems that make it easier for the user to organize and find email.

The user interface should allow the user to train the system in a simple, intuitive and transparent way by exploiting implicit and explicit feedback to the AI engine.

Required skills - Web development, Javascript
- Object oriented programming
- Relational databases, SQL
- Fundamentals of UI/UX
Skills should be obtained through passing the following courses with high scores:
- Web Applications I (and, optionally, Web Applications II)
- Data Science & Database Technology
- System and Device Programming
- Human Computer Interaction (preferably)


Deadline 10/03/2024      PROPONI LA TUA CANDIDATURA




© Politecnico di Torino
Corso Duca degli Abruzzi, 24 - 10129 Torino, ITALY
Contatti