An ESA implementation in python.
Project description
Get the required resources
scp -r webis@webislab40.medien.uni-weimar.de:/home/weci2587/projects/args-topic-modeling/resources .
To run the ESA-script with all terms run:
For normal ESA:
./esa-all-terms.py --similarity cos
--matrix-path <path_to_resources>/resources/esa-plain/<debatepedia|strategic-intelligence|wikipedia>.mat
--model-path <path_to_resources>/resources/esa-w2v/GoogleNews-vectors-negative300.bin
--model-vocab <path_to_resources>/resources/esa-w2v/w2v-vocab.p
--input-path <path_to_input_file>
--output-path <path_to_output_file>
For word2vec-ESA:
./esa-all-terms.py --similarity max
--matrix-path <path_to_resources>/resources/esa-w2v/<debatepedia|strategic-intelligence|wikipedia>.mat
--model-path <path_to_resources>/resources/esa-w2v/GoogleNews-vectors-negative300.bin
--model-vocab <path_to_resources>/resources/esa-w2v/w2v-vocab.p
--input-path <path_to_input_file>
--output-path <path_to_output_file>
To run the word2vec-ESA with reduced terms run:
./esa-top-n-terms.py -n <number_of_terms>
--corpus-path <path_to_resources>/resources/corpora/<debatepedia|strategic-intelligence|wikipedia>.csv
--model-path <path_to_resources>/resources/esa-w2v/GoogleNews-vectors-negative300.bin
--model-vocab <path_to_resources>/resources/esa-w2v/w2v-vocab.p
--input-path <path_to_input_file>
--output-path <path_to_output_file>
The input document must be a csv file with "|" as the separator and must contain the column "document", which is used as the input text for the ESA.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Close
Hashes for argument_esa_model-3.0.8.linux-x86_64.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 36bc05e193548e8179a12baa3b363a2ca79b2646276723c85f0e95e1b2dfba08 |
|
MD5 | f799b017096d9b843407857d07e72744 |
|
BLAKE2b-256 | a58137461e3fbc15abb1f522b7be8e111c70b3005dad3069af61ead76c1f1d3e |
Close
Hashes for argument_esa_model-3.0.8-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2c057499d016239b66128d402301e900bc831a4af15aad17e99ceb2f33264973 |
|
MD5 | 6322135b4bde98063346072c52ddd2ae |
|
BLAKE2b-256 | 8e289bfa9439c281f83a984368be5ca45ef0975556455ac7a9ba96d57cb3e9a4 |