#G2PT82 – 82º Geek Girls Portugal – Lisboa

Contamos contigo no dia 24 de Outubro a partir das 18h30 nas instalações da Talkdesk em Lisboa ?

O encontro é gratuito mas requer inscrição. Informamos também que, após vários pedidos, desta vez as apresentações serão ambas em inglês para dar-mos oportunidade a um público mais amplo 🙂


18:30 – Welcome & Check In Attendees

18:50 – Apresentação das G2PT & Talkdesk

19:00 – “Application of Data Science in E-commerce Fraud” – Ana Mendes

19:30 – Coffee Break (gentilmente cedido pela Talkdesk)

19:45 – “Digging into Portuguese Political Manifests using Text Mining” – Susana Quartin e Manuela Almeida

20:15 – Encerramento

Application of Data Science in E-commerce Fraud” – Ana Mendes

Bio: “A long, long time ago Ana received a PhD in Information Systems and Computer Engineering from Técnico, Lisboa. Since then she has been working on data focused teams at Sapo, Talkdesk and Farfetch, where she is currently a Data Science Engineer. She has been focusing on the topic of fraud and data science applied to fraud detection for the past two years. Ana likes to join running competitions and she is proud to have completed a few on the top 3… counting backwards”

Abstract: “It all began in 1994 when people could start buying from the internet. With that came the first e-commerce fraudsters, who were making purchases with false names and real, but stolen, credit card numbers. Merchants started to realise that Mickey Mouse and Bill Clinton were common names cropping up on their orders, so they started investigating their customer data.

This cat-and-mouse race has evolved, with fraudsters buying stacks of real credit cards numbers from the dark-web or creating fake sites to make phony sales. Companies have evolved too, by putting in place several mechanisms to avoid fraud-related losses and damages.

In this presentation I will delve into the fascinating world of fraud and show the application of data science as a viable strategy to detect and prevent fraudulent payments in e-commerce.”

Digging into Portuguese Political Manifests using Text Mining” – Susana Quartin e Manuela Almeida


Susana Quartin and Manuela Almeida are both Data Scientists at Talkdesk since 2017. Although they have the same academic background in Applied Mathematics, Manu enjoys mostly digging into data and working side-by-side with Product, while Susana prefers to wear the developer hat and build production solutions out of prototypes. Outside work, Manu loves to immerse herself in vegan baking (and eating it afterwards). Susana, on the other hand, is a couch potato that likes to watch dubious netflix series.


One of Talkdesk’s main lines of research in the Data Science area has been around topic extraction and analysis of calls’ transcriptions. Although we have faced many challenges that are specific to oral conversations, the methods we have used can be applied to other domains. In this talk we’ll share some of the results and insights we obtained from applying topic modeling and text summarization techniques (e.g., LDA and Sum Basic algorithm) to the Portuguese political parties’ legislative election manifests.



Talkdesk Logo