Search Results for author: Ivan Sekulić

Found 11 papers, 6 papers with code

Towards Self-Contained Answers: Entity-Based Answer Rewriting in Conversational Search

1 code implementation4 Mar 2024 Ivan Sekulić, Krisztian Balog, Fabio Crestani

One approach expands answers with inline definitions of salient entities, making the answer self-contained.

Conversational Search

Reliable LLM-based User Simulator for Task-Oriented Dialogue Systems

no code implementations20 Feb 2024 Ivan Sekulić, Silvia Terragni, Victor Guimarães, Nghia Khau, Bruna Guedes, Modestas Filipavicius, André Ferreira Manso, Roland Mathis

Notably, we have observed that fine-tuning enhances the simulator's coherence with user goals, effectively mitigating hallucinations -- a major source of inconsistencies in simulator responses.

Data Augmentation Task-Oriented Dialogue Systems +1

Estimating the Usefulness of Clarifying Questions and Answers for Conversational Search

no code implementations21 Jan 2024 Ivan Sekulić, Weronika Łajewska, Krisztian Balog, Fabio Crestani

While the body of research directed towards constructing and generating clarifying questions in mixed-initiative conversational search systems is vast, research aimed at processing and comprehending users' answers to such questions is scarce.

Conversational Search Retrieval

Evaluating Mixed-initiative Conversational Search Systems via User Simulation

1 code implementation17 Apr 2022 Ivan Sekulić, Mohammad Aliannejadi, Fabio Crestani

Clarifying the underlying user information need by asking clarifying questions is an important feature of modern conversational search system.

Conversational Search Text Generation +1

User Engagement Prediction for Clarification in Search

1 code implementation8 Feb 2021 Ivan Sekulić, Mohammad Aliannejadi, Fabio Crestani

Prompting the user for clarification in a search session can be very beneficial to the system as the user's explicit feedback helps the system improve retrieval massively.

Conversational Search Information Retrieval +1

Longformer for MS MARCO Document Re-ranking Task

1 code implementation20 Sep 2020 Ivan Sekulić, Amir Soleimani, Mohammad Aliannejadi, Fabio Crestani

Two step document ranking, where the initial retrieval is done by a classical information retrieval method, followed by neural re-ranking model, is the new standard.

Document Ranking Information Retrieval +2

Reasoning with Latent Structure Refinement for Document-Level Relation Extraction

2 code implementations ACL 2020 Guoshun Nan, Zhijiang Guo, Ivan Sekulić, Wei Lu

Document-level relation extraction requires integrating information within and across multiple sentences of a document and capturing complex interactions between inter-sentence entities.

Document-level Relation Extraction Relation +2

Not Just Depressed: Bipolar Disorder Prediction on Reddit

no code implementations12 Nov 2018 Ivan Sekulić, Matej Gjurković, Jan Šnajder

Bipolar disorder, an illness characterized by manic and depressive episodes, affects more than 60 million people worldwide.

Cannot find the paper you are looking for? You can Submit a new open access paper.