SEA: Entity search

In this edition of SEA we will discuss entity search. We have two amazing speakers lined up: Emma Gerritse (Radboud University Nijmegen) and Ridho Reinanda (Bloomberg)
This will be a hybrid event, the in-person event will take place at Lab42, Science Park, room L3.36.
***
IMPORTANT: You will be able to view the Zoom link once you 'attend' the meetup on this page.
***
17.00: Emma J Gerritse (Radboud University Nijmegen)
Title: Entity-aware Transformers for Entity Search
Abstract: Pre-trained language models such as BERT have been a key ingredient to achieve state-of-the-art results on a variety of tasks in natural language processing and, more recently, also in information retrieval. Recent research even claims that BERT is able to capture factual knowledge about entity relations and properties, the information that is commonly obtained from knowledge graphs. This talk investigates the following question: Do BERT-based entity retrieval models benefit from additional entity information stored in knowledge graphs? We show that the entity-enriched BERT model improves effectiveness on entity-oriented queries over a regular BERT model, establishing a new state-of-the-art result for the entity retrieval task, with substantial improvements for complex natural language queries and queries requesting a list of entities with a certain property. Additionally, we show that the entity information provided by our entity-enriched model particularly helps queries related to less popular entities. Last, we observe empirically that the entity-enriched BERT models enable fine-tuning on limited training data, which otherwise would not be feasible due to the known instabilities of BERT in few-sample fine-tuning, thereby contributing to data-efficient training of BERT for entity search.

17.30: Ridho Reinanda (Bloomberg)
Title: Entity-centric Tasks & Applications in Finance
Abstract: Data is the backbone of the financial markets, and one natural way of organizing such data is around entities, such as companies, securities, topics, and people. At Bloomberg, keeping track of the various information around entities is important in enabling several decision-support applications that we provide in the Bloomberg Terminal.
In this talk, we will introduce some of the attributes/relationships related to the entities that we model, as well as look at some of the specific challenges in the financial domain in terms of managing the lifecycle of entities. We will also go over a few use cases where entity information is leveraged in our applications, and discuss research our group has performed around entity-centric tasks.

SEA talks #236 and #237.