Scientific Research Project Title

KARE: Robust Exploratory Search in Large Knowledge Graphs

Research Institution

Aarhus University

Aarhus University

Research field

Databases, Data mining, Machine Learning

Research leader

Davide Mottin

Tenure Track Assistant Professor

Project title

KARE: Robust Exploratory Search in Large Knowledge Graphs

What is your project about?

In a world of interlinked individuals and systems, connections provide the necessary means to understand complex phenomena, such as the spread of disease. These connections form large networks called knowledge graphs. Such knowledge graphs require complicated analyses that traverse the network intelligently and return meaningful connections. Unfortunately, domain experts in relevant disciplines, such as medicine or social sciences, are generally not data-savvy nor data analysts. Consequently, the way these experts search the data is by exploration, resulting in vague search terms that return unsatisfactory answers and require several attempts to obtain the correct result. This search-and-refine process impedes fast discoveries in times in which immediacy is crucial. In addition to that, the search mechanism might return erroneous answers due to data imperfections.  This project targets both the vagueness and errors in exploratory search and proposes robust methods unsusceptible to data imperfections. To deliver robustness, the user provides the knowledge to guide intelligent algorithms to the correct answer. The project will deliver novel algorithmic methods for exploratory analyses of large networks that adjust the search to the user imprecisions and is robust towards the errors in such networks.

How did you become interested in your particular field of research?

I have been always fascinated by the possibility that computers one day could assist humans in their questions.  I saw the first search engines and smart-home assistants providing answers to simple questions using intelligent algorithms. As such, I decided to study ways to make information more accessible to anyone, whether data expert or not. This interest still drives my research.

What are the scientific challenges and perspectives in your project?

Two principal shortcomings in current research on graph search hinder easy exploration of the data: the inability to successfully resolve a vague search and the fragility of the search mechanism to errors in the data. As a result, the user has to modify the data or the search manually to reach an appropriate answer, thus delaying potential discoveries. This project proposes the first robust (i.e., unsusceptible to errors) exploration of large knowledge graphs. My project will result in modification algorithms to escort the user in the search process and cleaning algorithms robust to noise and errors in the data. These algorithms will drive smart-home assistants and specialized search engines with exploratory capabilities with which a user can easily find the requested information robustly.

What is your estimate of the impact, which your project may have to society in the long term?

Smart-home devices and specialized search engines are instrumental in searching and categorizing complex information. The KARE project will help anyone access such information robustly by returning the correct answer to a question and guiding the user in the search process. My project has an enormous potential to help prevent the spread of misinformation and speed up research in critical areas such as virology, physics, energy, in which promptness is a determinant factor for success.

Which impact do you expect the Sapere Aude programme will have on your career as a researcher?

This project is the coronation of several years of work in graph analysis, a discipline that requires contributions in the competitive areas of Databases, Data Mining, and Machine Learning. The proposed project will allow my development as an independent researcher and a world leader in graph analysis. Fighting misinformation and democratizing access to rich knowledge graphs will impact both society and the research. This project will help my professional growth and pave further interdisciplinary collaborations.

Background and personal life

I come from Marostica, a tiny village in Northern Italy at the feet of the Alps. I moved to Trento to complete my studies and my PhD; after that, I spent three years in Berlin as a postdoctoral researcher. I have moved to Aarhus in 2018 when I joined Aarhus University as a Tenure Track Assistant Professor. I like the Aarhusian atmosphere of a small city with a vibrant background. In my free time, I enjoy spending time with my wife, cooking, hiking, and swimming.

City of your current residence

Aarhus

High school

Liceo Scientifico Jacopo Da Ponte, Bassano del Grappa, Italy