Algoliterary Encounters: Difference between revisions
From Algolit
(→Creating word embeddings using word2vec) |
(→Creating word embeddings using word2vec) |
||
| Line 46: | Line 46: | ||
==== Creating word embeddings using word2vec ==== | ==== Creating word embeddings using word2vec ==== | ||
| − | * [[ | + | * [[Crowd Embeddings]] - case studies, still needs fine tuning |
* [[word2vec_basic.py]] - in piles of paper | * [[word2vec_basic.py]] - in piles of paper | ||
* [[softmax annotated]] | * [[softmax annotated]] | ||
Revision as of 12:57, 25 October 2017
Start of the Algoliterary Encounters catalog.
General Introduction
Algoliterary works
- Oulipo scripts
- i-could-have-written-that
- Obama, model for a politician
- In the company of CluebotNG
Algoliterary explorations
A few outputs to see how it works
- CHARNN text generator
- You shall know a word by the company it keeps - Five word2vec graphs, each of them containing the words 'collective', 'being' and 'social'.
Parts of NN process
Datasets
- Many many words - introduction to the datasets with calculation exercise
- The data (e)speaks - espeak installation
- The Enron email archive
- Common Crawl (used by GloVe): selection of urls (Constant, Maison du Livre...)
- Google News (used by word2vec)
- Frankenstein
- Learning from Deep Learning
- HG Wells personal dataset
- Jules Verne (FR), Shakespeare (FR) -> download from Gutenberg & clean up
- AnarchFem
- WikiHarass
- Tristes Tropiques
From words to numbers
Different views on the data
Creating word embeddings using word2vec
- Crowd Embeddings - case studies, still needs fine tuning
- word2vec_basic.py - in piles of paper
- softmax annotated
- Reverse Algebra
Autonomous machine as inspection
Algoliterary Toolkit
Bibliography
- Algoliterary Bibliography - Reading Room texts