Dynamic Topic Models of 'Dynamic Topic Modelling for Exploring the Scientific Literature on Coronavirus: An Unsupervised Labelling Technique'
Gespeichert in:
| Titel: | Dynamic Topic Models of 'Dynamic Topic Modelling for Exploring the Scientific Literature on Coronavirus: An Unsupervised Labelling Technique' |
|---|---|
| Autoren: | Guillén-Pacho, Ibai, orcid:0000-0001-7801- |
| Weitere Verfasser: | Badenes-Olmedo, Carlos, Corcho, Oscar |
| Verlagsinformationen: | Zenodo |
| Publikationsjahr: | 2024 |
| Bestand: | Zenodo |
| Schlagwörter: | Topic Models, Dynamic Topic Models, Dynamic Topic Labelling, Topic Labelling |
| Beschreibung: | This resource includes the models generated for the work Dynamic Topic Modelling for Exploring the Scientific Literature on Coronavirus: An Unsupervised Labelling Technique. Each zip file has the models with the different configurations (number of topics) for each type and, in addition, an evaluation script (bench.py) and different files necessary for this (localizer, timestamps, CORPUS etc.) are included. The requirements for reusing these models are as follows: Unzip all files and install the required packages ("requirements.txt" file). Download the precompiled DTM implementation of https://github.com/magsilva/dtm/tree/master/bin or compile manually the original implementation https://github.com/blei-lab/dtm Download the DTM wrapper from https://github.com/piskvorky/gensim/releases/tag/3.8.3 ("gensim-3.8.3/gensim/models/wrappers/dtmmodel.py"). Download the DETM python implementation of https://github.com/quynhneo/detm. To run model evaluation: modify the imports in the "bench.py" file to match the DETM and DTM models location and their full path (instructions in the file documentation). To repeat our topic study: follow the "notebook.ipynb" instructions. The overview of this resource is: RESOURCES├── BERTopic│ ├── BERTopic_100│ ├── BERTopic_100_probabilities.npy│ ├── BERTopic_100_topics│ ├── BERTopic_100_topic_words│ ├── BERTopic_200│ ├── BERTopic_200_probabilities.npy│ ├── BERTopic_200_topics│ ├── BERTopic_200_topic_words│ ├── BERTopic_300│ ├── BERTopic_300_probabilities.npy│ ├── BERTopic_300_topics│ ├── BERTopic_300_topic_words│ ├── BERTopic_400│ ├── BERTopic_400_probabilities.npy│ ├── BERTopic_400_topics│ └── BERTopic_400_topic_words│├── DETM│ ├── detm_deberta_model1 # 100 topics model │ ├── detm_deberta_model1_beta.mat│ ├── detm_deberta_model2 # 200 topics model │ ├── detm_deberta_model2_beta.mat│ ├── detm_word2vec_model1 # 100 topics model │ ├── detm_word2vec_model1_beta.mat│ ├── detm_word2vec_model2 # 200 topics model │ ├── detm_word2vec_model2_beta.mat│ └── min_df_3333│ └── .│├── DTM_ALL│ ├── ... |
| Publikationsart: | other/unknown material |
| Sprache: | English |
| ISSN: | 2364-4168 |
| Relation: | https://zenodo.org/records/12750327; oai:zenodo.org:12750327; https://doi.org/10.5281/zenodo.12750327 |
| DOI: | 10.5281/zenodo.12750327 |
| Verfügbarkeit: | https://doi.org/10.5281/zenodo.12750327 https://zenodo.org/records/12750327 |
| Rights: | Creative Commons Attribution 4.0 International ; cc-by-4.0 ; https://creativecommons.org/licenses/by/4.0/legalcode |
| Dokumentencode: | edsbas.B9B268CD |
| Datenbank: | BASE |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstract: | This resource includes the models generated for the work Dynamic Topic Modelling for Exploring the Scientific Literature on Coronavirus: An Unsupervised Labelling Technique. Each zip file has the models with the different configurations (number of topics) for each type and, in addition, an evaluation script (bench.py) and different files necessary for this (localizer, timestamps, CORPUS etc.) are included. The requirements for reusing these models are as follows: Unzip all files and install the required packages ("requirements.txt" file). Download the precompiled DTM implementation of https://github.com/magsilva/dtm/tree/master/bin or compile manually the original implementation https://github.com/blei-lab/dtm Download the DTM wrapper from https://github.com/piskvorky/gensim/releases/tag/3.8.3 ("gensim-3.8.3/gensim/models/wrappers/dtmmodel.py"). Download the DETM python implementation of https://github.com/quynhneo/detm. To run model evaluation: modify the imports in the "bench.py" file to match the DETM and DTM models location and their full path (instructions in the file documentation). To repeat our topic study: follow the "notebook.ipynb" instructions. The overview of this resource is: RESOURCES├── BERTopic│ ├── BERTopic_100│ ├── BERTopic_100_probabilities.npy│ ├── BERTopic_100_topics│ ├── BERTopic_100_topic_words│ ├── BERTopic_200│ ├── BERTopic_200_probabilities.npy│ ├── BERTopic_200_topics│ ├── BERTopic_200_topic_words│ ├── BERTopic_300│ ├── BERTopic_300_probabilities.npy│ ├── BERTopic_300_topics│ ├── BERTopic_300_topic_words│ ├── BERTopic_400│ ├── BERTopic_400_probabilities.npy│ ├── BERTopic_400_topics│ └── BERTopic_400_topic_words│├── DETM│ ├── detm_deberta_model1 # 100 topics model │ ├── detm_deberta_model1_beta.mat│ ├── detm_deberta_model2 # 200 topics model │ ├── detm_deberta_model2_beta.mat│ ├── detm_word2vec_model1 # 100 topics model │ ├── detm_word2vec_model1_beta.mat│ ├── detm_word2vec_model2 # 200 topics model │ ├── detm_word2vec_model2_beta.mat│ └── min_df_3333│ └── .│├── DTM_ALL│ ├── ... |
|---|---|
| ISSN: | 23644168 |
| DOI: | 10.5281/zenodo.12750327 |