eellak / gsoc2018-3gm Goto Github PK
View Code? Open in Web Editor NEW💫 Automated codification of Greek Legislation with NLP
Home Page: https://openlaws.ellak.gr/
License: GNU General Public License v3.0
💫 Automated codification of Greek Legislation with NLP
Home Page: https://openlaws.ellak.gr/
License: GNU General Public License v3.0
Add all issue types to the Internet Archive Collection. The workaround is very easy: You can use the fetcher script, download them from et.gr and upload them with scripts/ia-upload-all.sh
Estimated Time: 4 days
Hi, anytime I try to access to your project's link I get this error on my browser (Firefox)
Did Not Connect: Potential Security Issue
Firefox detected an issue and did not continue to openlaws.ellak.gr. The website is either misconfigured or your computer clock is set to the wrong time.
It’s likely the website’s certificate is expired, which prevents Firefox from connecting securely.
What can you do about it?
openlaws.ellak.gr has a security policy called HTTP Strict Transport Security (HSTS), which means that Firefox can only connect to it securely. You can’t add an exception to visit this site.
The issue is most likely with the website, and there is nothing you can do to resolve it. You can notify the website’s administrator about the problem.
Learn more…
Hope to solve this issue soon, thanks
Please configure 3gm.ellak.gr to work with the VM at 83.212.109.156 .
Add the ability to see the full history of a codified version of a law.
This requires creating a page where the user can see at the same time all changes made by all amending laws to a legal text.
This is not difficult in terms of the algorithm / code. The required code is almost complete.
The difficulty lies on the development of the right user interface that will facilitate user experince user and enable the better understanding of the applied changes to the end user.
Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
Describe the solution you'd like
A clear and concise description of what you want to happen.
Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.
Additional context
Add any other context or screenshots about the feature request here.
Translate Project to Greek.
Docker support would allow much easier installation and should be easier to be platform agnostic.
The current installation documentation not only have its own platform specific instructions but also has multiple links to another project's installation documentation, all of which has their own platform specific instructions.
Add the ability of projection of the incorporation of a draft law into an existing legislation.
A use case could be the following:
The user gives the draft of a law from a consultation at opengov.gr and can preview the changes it brings to the existing legislation.
Make the web application responsive (using Bootstrap)
Estimated Time : 4 days
A tool for comparison with ground truth is needed. The tool will include
Man applications of natural language processing and machine learning to text can benefit from a controlled lexicon of expert-selected terms (i.e., a dictionary). This is especially true of highly technical language, such as legal text. However, no open source and freely-available dictionaries of this nature have been available in Greek. Creating new legal dictionaries would greatly benefit 3gm but also automatic codification of Greek legal text in general.
Total time: 3 weeks
Developments in Greek politics during the last few years have mandated, large pieces of legislative texts to be published in the GGG at once, the most recent being the publication of the fourth memorandum. Effective summarization of laws therefore, becomes more and more useful.
I propose training a machine learning algorithm that can provide a comprehensive summary of each law or act in relation to its size.
It would be good to add the required functionality in order to export and display the titles of the articles
Train an NER (Named Entity Recognizer) special for Government Gazette Texts using the NLP library spaCy. The NER should be extended (do not override the pre-existing labels and do not add new labels if they are not needed since that will drop the model's accuracy).
The pre-existing Greek NER can be found here
Estimated Time: 1.5 month
Develop a dependency parser with neural networks.
Refers to: https://github.com/eellak/gsoc2018-3gm/wiki/Contributing-To-The-Project#training-statistical-models-with-custom-tag-maps
Estimate Time : 3 months
Use differential versioning packages such as Vermongo to save entries deltas.
Refers to: https://github.com/thiloplanz/v7files/wiki/Vermongo
and https://github.com/eellak/gsoc2018-3gm/wiki/Contributing-To-The-Project#differential-versioning-with-mongodb
Estimate Time : 2 weeks
Analyzing a corpus allows us to draw conclusions concerning its contents. Another way to understand how legislation is organized is to study closely the legislative graphs produced by the codifier.
Total time: 10 days
Reconstruct documents 1976-1998
Related blog post
Estimated Time: 1 week
Add the ability for interactive feedback / amendments on the algorithmically generated codified text by the end users.
In some cases, the resulting codified text contains erroneous references or is missing some references (links). For those cases , it would be nice to have a procedure that will allow:
In cases where more than one articles are being inserted into another Law, the system
retains by the numbering of that article the first imported, omits the others and considers Article 0 as the last one imported.
e.g. for Law 4512/2018 Article 353 which introduces new Articles 71A, 71B and 71C into another Law concludes as follows : https://3gm.ellak.gr/statute/l_4512_2018/codified
That is, the text of 71C was introduced as C.
Format routes to adhere to EUR-Lex Standards by the EU.
Example Page
Estimated Time: 7 days
Change UI as follows:
codify_law.html
history.html
Accordion elements
... (κείμενο των αλλαγών)
(Δεσμός:) [Εμφάνιση του ν. 4009/2011 μέχρι και τις αλλαγές του ν. 4485/2017]
Απόσπασμα Συνδέσμων + Status εφαρμογής.
... (κείμενο των αλλαγών)
(Δεσμός:) [Εμφάνιση του ν. 4009/2011 μέχρι και τις αλλαγές του ν. 4405/2016]
... (κείμενο των αλλαγών)
(Δεσμός:) [Εμφάνιση του ν. 4009/2011 μέχρι και τις αλλαγές του ν. 4310/2014]
The codifier module currently detects, codifies and stores all laws and presidential decrees that are found in the Greek Government Gazette issues.Even though these types of amendments are most important in Greek legislation they do not account for the largest part of GGG issues.
We want to broaden extraction capabilities using regular expressions to include parliamentary regulations (Κανονισμός της Βουλής), treaties (Συνθήκες), Prime Minister, Minister and Deputy Minister decisions, as well as acts of appointment, dismissal and transfer of public officials. An extensive account of types of legislative acts per GGG issue can be found in the [Ethnikon Typografeion website] (http://www.et.gr/index.php/f-e-k/teyxi).
As mentioned the legislative act extraction is currently performed using regular expressions in the entities module. An alternative to this is would be to train a neural network that can detect legislative acts and determines the type although this would be a complex solution depending on the number of issue types.
Since the GGG corpus contains a great number of types of legislative we should prioritize those found in main issue types such as 'Α', 'Β', 'Γ' , 'Δ'.
Broad range of fact extraction functionality, such as:
Work here should be done primarily using the re
built-in module.
You are as well free to use any tool (e.g. ML or something similar) to make it better
The code should be committed to entities.py
module:
Estimated Time : 2 weeks
sys.argv[1]
sys.stdin
sys.stdout
Example usage:
codifier.py ammendment-1.txt <initial-version.txt >ammended-version.txt
Optionally --input
and --output
flags
Extension
<initial-version.txt codifier.py ammendment-1.txt |
codifier.py ammendment-2.txt |
codifier.py ammendment-3.txt >final-version.txt
Add Apache Spark Support to MongoDB.
Use of ELI (https://publications.europa.eu/en/web/eu-vocabularies/eli) as a metadata & uri schema for the laws at 3gm.ellak.gr
Implement Help Page for explaining tags and functionality.
Check resources/phrases.txt
Develop an NER annotator module (in the form of a web app with flask) for annotating NER. One excellent (closed source sadly) is prodigy.ai
Estimated Time: 3 weeks
Estimated Time: 2-3 months
Do work on document and word embeddings. Refer to the Wiki Page for more information. We are using gensim
as a library.
Estimated Time: 1 week
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.