After the very warm reception of the first year, the fourth London Information Retrieval Meetup is approaching (11/02/2020) and we are excited to add more details about our speakers and talks!
The event is free and you are invited to register :

https://www.eventbrite.com/e/london-information-retrieval-meetup-february-2020-tickets-89056738101

Our first speaker is Anna Ruggero, one of our R&D software engineers:

Anna Ruggero

Anna Ruggero is a software engineer passionate about Information Retrieval and Data Mining.
She loves to find new solutions to problems, suggesting and testing new ideas, especially those that concern the integration of machine learning techniques into information retrieval systems.
Anna came into contact with search engines during her studies falling in love with this world, therefore she decided to investigate this topic further participating to the 12th European Summer School in Information Retrieval and doing her master degree dissertation on Entity Search.
Thanks to this path, she has expanded and improved her knowledges of Java and Python languages, information retrieval systems, clustering and word embeddings.

Entity Search on Virtual Documents Created with Graph Embeddings

Entity Search is a search paradigm that aims to retrieve entities and all the information related to them. In the last few years the importance of this topic has become greater and greater due to the fact that 40% of the queries made by users mention specific entities nowdays.
This talk wants to give a first overview of the state-of-the-art methods used for entities retrieval and then describe the new approach Anna has implemented and proposed in her master thesis. The novelty introduced with this work exploits two machine learning techniques: neural network and clustering.
In particular:
– Graph embeddings were used for the creation of the entities representations.
– Clustering was used for the creation of those virtual documents necessary for the retrieval phase.
The aim of this work is to integrate existing machine learning techniques with traditional Information Retrieval, expanding the concept of document to a set of related and highly connected entities.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.