Merge branch 'arxmliv-2018' into 'master'

arxmliv 08.2018 release See merge request !2

Merge branch 'arxmliv-2018' into 'master'
cf10c1a0 · Deyan Ginev · 8aba14b2 · e2eb4cf1 · cf10c1a0 · cf10c1a0
Commit cf10c1a0 authored 6 years ago by Deyan Ginev
--- a/resources/arxmliv-dataset-082018.md
+++ b/resources/arxmliv-dataset-082018.md
+---
+layout: page
+title: arXMLiv 08.2018 - An HTML5 dataset for arXiv.org 
+---
+Part of the [arXMLiv](https://kwarc.info/systems/arXMLiv/) project at the [KWARC](https://kwarc.info/) research group
+### Author
+ - Deyan Ginev
+### Current release
+ - 08.2018
+### Accessibility and License
+The content of this Dataset is licensed to [SIGMathLing members](/member/) for research
+and tool development purposes. 
+Access is restricted to  [SIGMathLing members](/member/) under the
+[SIGMathLing Non-Disclosure-Agreement](/nda/) as for most [arXiv](http://arxiv.org)
+articles, the right of distribution was only given (or assumed) to arXiv itself.
+### Contents
+  - 1,232,186 HTML5 documents
+  - Three separate archive bundles separated by LaTeXML conversion severity
+  - derivative **word embeddings** and a **token model** are available separately [here](/resources/arxmliv-embeddings-082018/)
+  | subset ID  | number of documents | size archived | size unpacked |
+  | :---       | ---:                | ---:          | ---:          |
+  | no\_problem| 137,864             | 6.9 GB        | 53 GB         |
+  | warning    | 705,095             | 104 GB        | 896 GB        | 
+  | error      | 389,227             | 58 GB         | 506 GB        |
+  | subset file name                 | MD5                                 |
+  | :---                             | :---                                |
+  | `arXMLiv_08_2018_no_problem.zip` | `0e19fd0a68f18f6bb05b3d7d1127a072`  |
+  | `arXMLiv_08_2018_warning.zip`    | `8ea32275e9cf197f8eb592d539206a6f`  | 
+  | `arXMLiv_08_2018_error.zip`      | `5880d4fac417f5fdad6e8cec0e3bfe2a`  |
+**Note:** the archives preserve a leading path `./data/datasets/dataset-arXMLiv-08-2018`, which can be safely ignored, and is an artefact of the current release that will be avoided in the future.
+### Description
+This is a second public release of the arXMLiv dataset generated by the [KWARC](https://kwarc.info/) research group. It contains 1,232,186 HTML5 scientific documents from the arXiv.org preprint archive, converted from their respective TeX sources. A 13% increase in available articles over the 08.2017 release.
+The dataset is segmented in 3 different subsets, each corresponding to a severity level of the LaTeXML software responsible for the HTML5 conversion. 
+ - The `no_problem` set had no obvious challenges in conversion and is the safest, most reliable subset
+ - The `warning` set covers a variety of minor issues, from mathematical expressions unparseable by the LaTeXML grammar, to missing LaTeX packages with no apparent use in the document. The vast majority of the documents should both have a good-looking rendering, as well as data consistency for e.g. NLP tasks.
+ - The `error` set covers all conversions which successfully generated an HTML5 document, but had major issues during the conversion. Examples would range from unknown macros (due to limited LaTeX coverage), unexpected latex syntax, math/text mode mismatches, as well as real LaTeX errors from the original sources. This subset should be used with extra caution, though should still preserve overall data consistency and could be safely used for e.g. generating word embeddings.
+ This version of the dataset has had minimal manual quality control, and we offer no additional warranty beyond the latexml severity reported.
+We welcome community feedback on all of: data quality, representation issues, need for auxiliary resources (e.g. figures, token models), as well as organization and archival best practices. The conversion, build system, and data redistribution efforts are all ongoing projects at the [KWARC research group](http://kwarc.info).
+### Citing this Resource
+The dataset should be referenced in all academic publications that present results
+obtained with its help. The reference should contain the identifier `arXMLiv:08.2018` in
+the title, the author, year, a reference to SIGMathLing, and the URL of the resource
+description page. For convenience, we supply some records for bibTeX and EndNote below. To
+cite a particular part of the dataset use the subset identifiers in the ciation; 
+e.g. `\cite[no_problem subset]{arXMLiv:08.2018}` or just explain it in the text using the
+concrete identifier.  
+#### pure bibTeX
+```
+@MISC{SML:arXMLiv:08.2018,
+  author = {Deyan Ginev},
+  title = {arXMLiv:08.2018 dataset, an HTML5 conversion of arXiv.org},
+  howpublished = {hosted at \url{https://sigmathling.kwarc.info/resources/arxmliv/}},
+  note = {SIGMathLing -- Special Interest Group on Math Linguistics},
+  year = 2018}
+```
+#### bibTeX for the bibLaTeX package (preferred)
+```
+@online{SML:arXMLiv:08.2018,
+  author = {Deyan Ginev},
+  title = {arXMLiv:08.2018 dataset, an HTML5 conversion of arXiv.org},
+  url = {https://sigmathling.kwarc.info/resources/arxmliv/},
+  note = {SIGMathLing -- Special Interest Group on Math Linguistics},
+  year = 2018}
+```
+#### EndNote
+```
+%0 Generic
+%T arXMLiv:08.2018 dataset, an HTML5 conversion of arXiv.org
+%A Ginev, Deyan
+%D 2018
+%I hosted at https://sigmathling.kwarc.info/resources/arxmliv/
+%F SML:arXMLiv:08.2018b
+%O SIGMathLing – Special Interest Group on Math Linguistics
+```
+### Download
+  [Download link](https://gl.kwarc.info/SIGMathLing/dataset-arXMLiv-08-2018)
+  ([SIGMathLing members](/member/) only)
+### Generated via
+ - [LaTeXML 0.8.3](https://github.com/brucemiller/LaTeXML/releases/tag/v0.8.3), 
+ - [CorTeX 0.3.1](https://github.com/dginev/CorTeX/releases/tag/0.3.1)
--- a/resources/arxmliv-embeddings-082018.md
+++ b/resources/arxmliv-embeddings-082018.md
+---
+layout: page
+title: arXMLiv 08.2018 - Word Embeddings; Token Model
+---
+Part of the [arXMLiv](https://kwarc.info/systems/arXMLiv/) project at the [KWARC](https://kwarc.info/) research group
+### Author
+ - Deyan Ginev
+### Current release
+ - 08.2018
+### Accessibility and License
+The content of this Dataset is licensed to [SIGMathLing members](/member/) for research
+and tool development purposes. 
+Access is restricted to  [SIGMathLing members](/member/) under the
+[SIGMathLing Non-Disclosure-Agreement](/nda/) as for most [arXiv](http://arxiv.org)
+articles, the right of distribution was only given (or assumed) to arXiv itself.
+### Contents
+  - An 11.5 billion token model for the arXMLiv 08.2018 dataset, including subformula lexemes
+    - `token_model.zip`
+  - 300 dimensional GloVe word embeddings for the arXMLiv 08.2018 dataset
+    - `glove.arxmliv.5B.300d.zip` and `vocab.arxmliv.zip`
+  - 300d GloVe word embeddings for individual subsets
+    - `glove.subsets.zip`
+  - the main arXMLiv dataset is available separately [here](/resources/arxmliv-dataset-082018/)
+#### Token Model Statistics
+subset     | documents | paragraphs | sentences   |
+-----------|----------:|-----------:|------------:|
+no_problem | 137,864   | 4,646,203  | 21,533,963  |
+warning    | 705,095   | 45,797,794 | 183,246,777 |
+error      | 389,225   | 26,759,524 | 99,641,978  |
+complete   | 1,232,184 | 77,203,521 | 304,422,718 |
+subset     | words         | formulas    | inline cite | numeric literals |
+-----------|--------------:|-----------: |------------:|-----------------:|
+no_problem | 430,217,995   | 20,910,732  | 3,709,520   | 11,177,753       |
+warning    | 3,175,663,430 | 281,832,412 | 25,337,574  | 83,606,897       |
+error      | 1,731,971,035 | 153,186,264 | 13,145,561  | 43,399,720       |
+complete   | 5,337,852,460 | 455,929,408 | 42,192,655  | 138,184,370      |
+#### GloVe Model Statistics
+subset     | tokens         | unique words | unique words (freq 5+ )
+-----------|--------------: |-------------:|-----------------------:
+no_problem | 622,968,267    | 715,433      | 219,304
+warning    | 7,203,536,205  | 3,478,235    | 666,317
+error      | 3,691,805,321  | 2,444,532    | 574,467
+complete   | 11,518,309,793 | 5,285,379    | 1,000,295
+### Citing this Resource
+Please cite the main dataset when using the word embeddings, as they are generated and distributed jointly. [Instructions here](/resources/arxmliv-dataset-082018/#citing-this-resource)
+### Download
+  [Download link](https://gl.kwarc.info/SIGMathLing/embeddings-arXMLiv-08-2018)
+  ([SIGMathLing members](/member/) only)
+### Generated via
+ - [llamapun 0.2.0](https://github.com/KWARC/llamapun/releases/tag/0.2.0), 
+ - [GloVe 1.2, 2018](https://github.com/stanfordnlp/GloVe/tree/07d59d5e6584e27ec758080bba8b51fce30f69d8)
+### Generation Parameters
+ * token model distributed as 3 subsets - no_problem, warning and error. complete model is derived via:
+    ```
+      cat token_model_no_problem.txt \
+          token_model_warning.txt \
+          token_model_error.txt > token_model_complete.txt
+    ```
+ * [llamapun v0.2](https://github.com/KWARC/llamapun/releases/tag/0.2), `corpus_token_model` example used for token model extraction
+   * processed logical paragraphs only (excluded non-textual modalities, e.g. tables, figures and their captions, bibliographies, as well as abstracts)
+   * excluded paragraphs containing latexml errors (marked via a `ltx_ERROR` HTML class)
+   * used llamapun math-aware sentence and word tokenization, with subformula math lexemes
+   * marked up inline citations replaced with `citationelement` token
+   * numeric literals replaced with `NUM` token (both in text and formulas)
+   * ignored sentences with unnaturally long words (>25 characters) - almost always due to latexml conversion errors
+ * [GloVe repository at sha 07d59d](https://github.com/stanfordnlp/GloVe/tree/07d59d5e6584e27ec758080bba8b51fce30f69d8)
+   * build/vocab_count -min-count 5
+   * build/cooccur -memory 32.0 -window-size 15
+   * build/shuffle -memory 32.0
+   * build/glove -threads 16 -x-max 100 -iter 25 -vector-size 300 -binary 2
+### Examples and baselines
+#### GloVe in-built evaluation (non-expert tasks e.g. language, relationships, geography)
+1. no_problem
+  * Total accuracy: 27.47%  (4028/14663)
+  * Highest score: "gram3-comparative.txt", 75.30% (1003/1332)
+2. warning
+  * Total accuracy: 31.97%  (5351/16736)
+  * Highest score: "gram3-comparative.txt", 77.40% (1031/1332)
+3. error
+  * Total accuracy: 29.67%  (4910/16549)
+  * Highest score: "gram3-comparative.txt", 72.75% (969/1332)
+4. complete
+  * Total accuracy: 35.48%  (6298/17750) 
+  * Highest score: "gram3-comparative.txt", 76.65% (1021/1332)
+5. demo baseline: text8 demo (first 100M characters of Wikipedia)
+  * Total accuracy: 23.62%  (4211/17827)
+  * Highest score: "gram6-nationality-adjective.txt", 58.65% (892/1521)
+**Evaluation note:** These in-built evlauation runs are provided as a sanity check that the generated GloVe models pass a basic baseline against the non-expert tasks in the default GloVe suite. 
+One would need a scienctific discourse tailored set of test cases to evaluate the arXiv-based models competitively.
+#### Measuring word analogy
+In a cloned GloVe repository, start via:
+```
+python2 eval/python/word_analogy.py --vocab_file vocab.arxmliv.txt --vectors_file glove.arxmliv.11B.300d.txt 
+```
+1. `abelian` is to `group` as `disjoint` is to `?`
+  * Top hit: `union`, cosine distance	`0.653377`
+2. `convex` is to `concave` as `positive` is to `?`
+  * Top hit: `negative`, cosine distance `0.786877`
+3. `finite` is to `infinte` as `abelian` is to `?`
+  * Top hit: `nonabelian`, cosine distance `0.676042`
+4. `quantum` is to `classical` as `bottom` is to `?`
+  * Top hit: `top`, cosine distance `0.734896`
+5. `eq` is to `proves` as `figure` is to `?`
+  * Top hit: `showing`, cosine distance `0.668502`
+6. `italic_x` is to `italic_y` as `italic_a` is to `?`
+  * Top hit: `italic_b`, cosine distance `0.902608`
+#### Nearest word vectors 
+In a cloned GloVe repository, start via:
+```
+python2 eval/python/distance.py --vocab_file vocab.arxmliv.txt --vectors_file glove.arxmliv.11B.300d.txt 
+```
+1. **lattice** 
+```
+Word: lattice  Position in vocabulary: 488
+                               Word       Cosine distance
+---------------------------------------------------------
+                           lattices		0.853103
+                         triangular		0.637767
+                          honeycomb		0.626426
+                            crystal		0.624397
+                             finite		0.614720
+                            spacing		0.603067
+```
+2. **entanglement**
+```
+Word: entanglement  Position in vocabulary: 1568
+                               Word       Cosine distance
+---------------------------------------------------------
+                          entangled		0.780425
+                       multipartite		0.730968
+                        concurrence		0.691708
+                         negativity		0.649595
+                         tripartite		0.647623
+                            quantum		0.640395
+                           fidelity		0.640285
+                      teleportation		0.616797
+                            discord		0.613752
+                            entropy		0.612341
+                          bipartite		0.608034
+                          coherence		0.606859
+                        nonlocality		0.601337
+```
+3. **forgetful**
+```
+Word: forgetful  Position in vocabulary: 11740
+                               Word       Cosine distance
+---------------------------------------------------------
+                            functor		0.723472
+                           functors		0.656184
+                           morphism		0.598965
+```
+4. **eigenvalue**
+```
+Word: eigenvalue  Position in vocabulary: 1448
+                               Word       Cosine distance
+---------------------------------------------------------
+                        eigenvalues		0.893073
+                        eigenvector		0.768380
+                       eigenvectors		0.765241
+                      eigenfunction		0.754222
+                     eigenfunctions		0.686141
+                         eigenspace		0.666098
+                              eigen		0.641422
+                             matrix		0.616723
+                          eigenmode		0.613117
+                         eigenstate		0.612188
+                          laplacian		0.611396
+                            largest		0.606122
+                           smallest		0.605342
+                         eigenmodes		0.604839
+```
+5. **riemannian**
+```
+Word: riemannian  Position in vocabulary: 2285
+                               Word       Cosine distance
+---------------------------------------------------------
+                          manifolds		0.765827
+                           manifold		0.760806
+                             metric		0.719817
+                            finsler		0.687826
+                          curvature		0.676100
+                              ricci		0.664770
+                            metrics		0.660804
+                         riemmanian		0.651666
+                          euclidean		0.644686
+                         noncompact		0.643878
+                        conformally		0.638984
+                          riemanian		0.633814
+                             kahler		0.632680
+                            endowed		0.622035
+                        submanifold		0.613868
+                       submanifolds		0.612716
+                           geodesic		0.604488
+```
\ No newline at end of file
--- a/resources/index.md
+++ b/resources/index.md
@@ -2,10 +2,10 @@
 layout: page
 title: SIGMathLing - Datasets and Resources
 ---
 1. [arXMLiv corpus, 08.2017 release](/resources/arxmliv-dataset-082017/)
 2. [arXMLiv word embeddings, 08.2017 release](/resources/arxmliv-embeddings-082017)
 3. [quantity expressions](/resources/quantity-expressions)
+ 4. [arXMLiv corpus, 08.2018 release](/resources/arxmliv-dataset-082018/)
+ 5. [arXMLiv word embeddings, 08.2018 release](/resources/arxmliv-embeddings-082018)
 Additional resources are en route, see the [plan](/technical/) for details.