Pinned Repositories
embedding_downsampling_comparison
Experiments on word embedding reliability with different sampling scenarios. Includes modified version of word2vec.
hellrich_coling2016
Code for experiments in Hellrich & Hahn: "Bad Company - Neighborhoods in Neural Embedding Spaces Considered Harmful", COLING 2016.
hellrich_dh2016
Johannes Hellrich & Udo Hahn (2016): Measuring the Dynamics of Lexico-Semantic Change Since the German Romantic Period. Digital Humanities 2016
hellrich_dh2017
Johannes Hellrich & Udo Hahn (2017): Don't Get Fooled by Word Embeddings-Better Watch their Neighborhood. Digital Humanities 2017.
hellrich_latech2016
Johannes Hellrich & Udo Hahn (2016): An Assessment of Experimental Protocols for Tracing Changes in Word Semantics Relative to Accuracy and Reliability. LaTeCH @ ACL 2016.
HistEmo
hyperwords
Modified version of Omar Levy's hyperwords word embedding tool, allows for weighted downsampling as well as resource friendly training of models.
JeSemE
The Jena Semantic Explorer website (and processing pipeline), one of the main contributions in my Ph.D. thesis. See also Hellrich et al.: "JeSemE: A Website for Exploring Diachronic Changes in Word Meaning and Emotion", COLING 2018
metal_lyric_topic_modeling
Code and topic models for experiments in Hellrich & Rzymski (2019): Computational Detection of Medieval References in Metal. In: Barratt-Peacock & Hagen (Eds.): Medievalism and Metal Music Studies: Throwing Down the Gauntlet. Emerald.
precision_medicine_dl_classifier
Document classification experiments on TREC precision medicine data. Adaption of a state-of-the-art deep learning algorithm with additional structured features.
hellrich's Repositories
hellrich/precision_medicine_dl_classifier
Document classification experiments on TREC precision medicine data. Adaption of a state-of-the-art deep learning algorithm with additional structured features.
hellrich/hyperwords
Modified version of Omar Levy's hyperwords word embedding tool, allows for weighted downsampling as well as resource friendly training of models.
hellrich/JeSemE
The Jena Semantic Explorer website (and processing pipeline), one of the main contributions in my Ph.D. thesis. See also Hellrich et al.: "JeSemE: A Website for Exploring Diachronic Changes in Word Meaning and Emotion", COLING 2018
hellrich/embedding_downsampling_comparison
Experiments on word embedding reliability with different sampling scenarios. Includes modified version of word2vec.
hellrich/hellrich_coling2016
Code for experiments in Hellrich & Hahn: "Bad Company - Neighborhoods in Neural Embedding Spaces Considered Harmful", COLING 2016.
hellrich/hellrich_dh2016
Johannes Hellrich & Udo Hahn (2016): Measuring the Dynamics of Lexico-Semantic Change Since the German Romantic Period. Digital Humanities 2016
hellrich/hellrich_dh2017
Johannes Hellrich & Udo Hahn (2017): Don't Get Fooled by Word Embeddings-Better Watch their Neighborhood. Digital Humanities 2017.
hellrich/hellrich_latech2016
Johannes Hellrich & Udo Hahn (2016): An Assessment of Experimental Protocols for Tracing Changes in Word Semantics Relative to Accuracy and Reliability. LaTeCH @ ACL 2016.
hellrich/HistEmo
hellrich/metal_lyric_topic_modeling
Code and topic models for experiments in Hellrich & Rzymski (2019): Computational Detection of Medieval References in Metal. In: Barratt-Peacock & Hagen (Eds.): Medievalism and Metal Music Studies: Throwing Down the Gauntlet. Emerald.
hellrich/jufit
Jena UMLS Filter - filters and enhances UMLS Files
hellrich/Mallet
MALLET is a Java-based package for statistical natural language processing, document classification, clustering, topic modeling, information extraction, and other machine learning applications to text.
hellrich/sgns_implementation_comparison
Comparing SGNS word embedding implementations