Papers by Anne-Kathrin Schumann
Training a Broad-Coverage German Sentiment Classification Model for Dialog Systems (2020.lrec-1)
Copied to clipboard
| Challenge: | Existing sentiment data sets are not available for sentiment analysis. |
| Approach: | They propose to combine a German sentiment corpus with existing resources to train a general-purpose German sentiment classification model. |
| Outcome: | The proposed model trains a general-purpose German sentiment classification model . the data set contains 5.4 million labelled samples . |
Automatic Annotation of Semantic Term Types in the Complete ACL Anthology Reference Corpus (L18-1)
Copied to clipboard
| Challenge: | a recent increase in quantitative studies of scientific text collections has led to a significant increase in the use of semantic labeling techniques. |
| Approach: | They propose to use semantic class labels to enhance a well-known resource . they use semantic labels to assign semantic class labeling to technical terms . |
| Outcome: | The proposed approach enhances the ACL Anthology Reference Corpus with semantic class labels for 20,000 technical terms . the goal is to use this information as one feature in the profiling of scientific papers, communities, and disciplines. |