Doctoral Candidate at NEWTRAL

rafael_high_resolution

Rafael Aparecido Martins Frade

DC 9 at NEWTRAL

Phd candidate at Newtral researching automated fact-checking and multi-modal machine learning.

I am currently researching automated fact-checking, namely multi-modal claim matching. The objective is to develop models and mechanisms to help fight the spread of fake news. Some of the applications we’re planning to work on are euroscepticism, climate emergency and health.
I have a bachelor’s degree in Social Sciences and a master’s in Economics. Before joining Newtral, I have worked as a software developer and as research assistant in projects of applied econometrics and economic growth. I am open to collaborations in projects involving automated fact-checking, semantic similarity, machine reasoning, information retrieval, multi-modal ML, applications of LLMs and machine learning applied to social sciences in general.

DC9 Individual Project

Claim verification for fact-checking

Research Objectives:

Enhancing Political Claim Verification with Contextual Analysis

To design a strategy to verify political claims, the verification consisting not only into assigning false or true to each claim, but also in supplying more context so as to get a more informative and actionable response.

Sourcing Reliable Information for Political Claim Verification

Find reliable sources of information to verify the claim (or factual assertion).

Semantic Similarity Assessment for Political Claims

Design a strategy to evaluate claim similarity based not on the sentence structure but on the basis of its semantic meaning and context.

Measuring Semantic Equivalence in Distinct Political Claims

Define a metric to evaluate if two different claims have the same meaning.

Resilient Algorithms for Fact-Checking Noisy Transcripts

Develop contextual algorithms capable of evaluating claim veracity based on claims that have been previously fact-checked by humans and develop techniques resilient to "noisy transcripts".

Expected Results:

Real-Time Veracity Checker: A Prototype for Live Political Fact-Checking

Development of a prototype to automatically check the veracity of political claims in real-time. This kind of prototypes would be of great help of live fact-checking of political debates.

Reliable Sources and Tools for Automated Political Fact-Checking

A list of reliable sources for political fact checking and automatic tools to look up their content.

Fact-Checking in Action: Use Cases on Climate Emergency, Immigration, and Euroscepticism

Use cases on climate emergency, immigration, and Euroscepticism.

Supervisors

In our pursuit of academic excellence, HYBRIDS Doctoral Candidates are guided by a dedicated team of supervisors. Comprising the Main Supervisor, Co-Supervisor, and Inter-sectoral Supervisor, this team of professionals offers a wealth of knowledge, mentorship, and interdisciplinary insights.

Main Supervisor

Dr. Rubén Míguez

Newtral Media Audiovisual (NEWTRAL)

Co-Supervisor

Dr. Arkaitz Zubiaga

Queen Mary University (QMUL)

Inter-sectoral Supervisors

Ms. Stephanie Öttl

Industrieanlagen- Betriebsgesellschaft (IABG)

Get in touch with us