Papers by Gerhard Heyer

8 papers
Casting the Same Sentiment Classification Problem (2021.findings-emnlp)

Copied to clipboard

Challenge: Identifying the stance of an argument towards a topic is a fundamental problem in computational argumentation.
Approach: They propose a task where text users are asked to determine if they have the same sentiment . they aim to enable a more topic-agnostic sentiment classification by using Yelp data .
Outcome: The proposed task achieves an accuracy above 83% for category subsets across topics and 89% on average.
On Classifying whether Two Texts are on the Same Side of an Argument (2021.emnlp-main)

Copied to clipboard

Challenge: Existing approaches to same side stance classification (S3C) require domain knowledge and semantic inference to solve the task.
Approach: They propose to use same side stance classification to predict whether two arguments argue for the same stance for a given pair of arguments.
Outcome: The proposed model fails to generalize both within and across topics and domains when adjusting the sampling strategy to a more adversarial scenario.
EVALIGN: Visual Evaluation of Translation Alignment Models (2023.eacl-demo)

Copied to clipboard

Challenge: EvAlign is a visual analytics framework for quantitative and qualitative evaluation of automatic translation alignment models.
Approach: They propose to use EvAlign to analyze automatic translation alignment models and compare their performance with other baseline and state-of-the-art models.
Outcome: The framework hosts nine gold standard datasets and the predictions of multiple alignment models.
ILCM - A Virtual Research Infrastructure for Large-Scale Qualitative Data (L18-1)

Copied to clipboard

Challenge: iLCM project develops integrated research environment for qualitative data analysis . text mining and text mining tools are extended by "Open Research Computing"
Approach: iLCM project develops integrated research environment for analysis of structured and unstructured data in a "Software as a Service" architecture.
Outcome: iLCM project develops integrated research environment for analysis of structured and unstructured data in a "Software as a Service" architecture.
Self-Training for Sample-Efficient Active Learning for Text Classification with Pre-Trained Language Models (2024.emnlp-main)

Copied to clipboard

Challenge: Existing methods to train models without labeled data are lacking in supervised tasks . a lack of labeles is the main obstacle to real-world applications .
Approach: They propose a semi-supervised approach that uses a model to obtain pseudo-labels for unlabeled data.
Outcome: The proposed method outperforms the reproduced methods on four text classification benchmarks.
Page Stream Segmentation with Convolutional Neural Nets Combining Textual and Visual Features (L18-1)

Copied to clipboard

Challenge: (retro-)digitizing paper-based files is a major undertaking for private and public archives and an important task in electronic mailroom applications.
Approach: They propose to use convolutional neural networks to combine image and text features to achieve optimal document separation.
Outcome: The proposed approach achieves an accuracy of 93 % and is considered a state-of-the-art for this task.
Press Freedom Monitor: Detection of Reported Press and Media Freedom Violations in Twitter and News Articles (2021.emnlp-demo)

Copied to clipboard

Challenge: Press freedom is under constant and increasing attack, even in Europe.
Approach: They introduce a tool that aims to detect reported press and media freedom violations in news articles and tweets.
Outcome: The tool has shown an impressive performance in detecting press and media freedom violations in news articles and tweets.
Supporting Land Reuse of Former Open Pit Mining Sites using Text Classification and Active Learning (2021.acl-long)

Copied to clipboard

Challenge: open pit mines left many regions worldwide inhospitable or uninhabitable . aforementioned information has to be acquired to ensure safety and validity of land reuse .
Approach: They propose a workflow for supporting the post-mining management of former open pit mines in the eastern part of Germany . they use active learning to perform multi-label sentence classification for two categories of restrictions and seven categories of topics .
Outcome: The proposed system supports the post-mining management of former lignite open pit mines in the eastern part of Germany.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations