Search Results for author: Julian Risch

Found 18 papers, 11 papers with code

Fabricator: An Open Source Toolkit for Generating Labeled Training Data with Teacher LLMs

1 code implementation18 Sep 2023 Jonas Golde, Patrick Haller, Felix Hamborg, Julian Risch, Alan Akbik

Here, a powerful LLM is prompted with a task description to generate labeled data that can be used to train a downstream NLP model.

Question Answering text-classification +2

PatentMatch: A Dataset for Matching Patent Claims & Prior Art

3 code implementations27 Dec 2020 Julian Risch, Nicolas Alder, Christoph Hewel, Ralf Krestel

For these reasons, we address the computer-assisted search for prior art by creating a training dataset for supervised machine learning called PatentMatch.

Information Retrieval Retrieval +1

Offensive Language Detection Explained

1 code implementation LREC 2020 Julian Risch, Robin Ruff, Ralf Krestel

However, even with machine-learned models achieving better classification accuracy than human experts, there is still a reason why human moderators are preferred.

BIG-bench Machine Learning General Classification +1

Bagging BERT Models for Robust Aggression Identification

1 code implementation LREC 2020 Julian Risch, Ralf Krestel

In this paper, we describe such an ensemble system and present our submission to the shared tasks on aggression identification 2020 (team name: Julian).

Aggression Identification text-classification +1

Top Comment or Flop Comment? Predicting and Explaining User Engagement in Online News Discussions

1 code implementation26 Mar 2020 Julian Risch, Ralf Krestel

In this paper, we systematically analyze user engagement in the form of the upvotes and replies that a comment receives.

Multifaceted Domain-Specific Document Embeddings

1 code implementation NAACL 2021 Julian Risch, Philipp Hager, Ralf Krestel

Current document embeddings require large training corpora but fail to learn high-quality representations when confronted with a small number of domain-specific documents and rare terms.

Document Embedding Knowledge Graphs

Prediction for the Newsroom: Which Articles Will Get the Most Comments?

1 code implementation NAACL 2018 Carl Ambroselli, Julian Risch, Ralf Krestel, Andreas Loos

The overwhelming success of the Web and mobile technologies has enabled millions to share their opinions publicly at any time.

regression

A Dataset of Journalists' Interactions with Their Readership: When Should Article Authors Reply to Reader Comments?

1 code implementation19 Oct 2020 Julian Risch, Ralf Krestel

Based on this data, we formulate the novel task of recommending reader comments to journalists that are worth reading or replying to, i. e., ranking comments in such a way that the top comments are most likely to require the journalists' reaction.

My Approach = Your Apparatus? Entropy-Based Topic Modeling on Multiple Domain-Specific Text Collections

1 code implementation25 Nov 2019 Julian Risch, Ralf Krestel

Comparative text mining extends from genre analysis and political bias detection to the revelation of cultural and geographic differences, through to the search for prior art across patents and scientific papers.

Bias Detection Clustering +2

Delete or not Delete? Semi-Automatic Comment Moderation for the Newsroom

no code implementations COLING 2018 Julian Risch, Ralf Krestel

Comment sections of online news providers have enabled millions to share and discuss their opinions on news topics.

Multi-modal Retrieval of Tables and Texts Using Tri-encoder Models

no code implementations EMNLP (MRQA) 2021 Bogdan Kostić, Julian Risch, Timo Möller

In this paper, we present an approach for retrieving both texts and tables relevant to a question by jointly encoding texts, tables and questions into a single vector space.

Extractive Question-Answering Question Answering +1

Semantic Answer Similarity for Evaluating Question Answering Models

no code implementations EMNLP (MRQA) 2021 Julian Risch, Timo Möller, Julian Gutsch, Malte Pietsch

To this end, we create an English and a German three-way annotated evaluation dataset containing pairs of answers along with human judgment of their semantic similarity, which we release along with an implementation of the SAS metric and the experiments.

Question Answering Semantic Similarity +1

Pseudo-Labels Are All You Need

no code implementations GermEval 2022 Bogdan Kostić, Mathis Lucka, Julian Risch

Automatically estimating the complexity of texts for readers has a variety of applications, such as recommending texts with an appropriate complexity level to language learners or supporting the evaluation of text simplification approaches.

Feature Engineering Pseudo Label +2

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