Constitutional awareness dedicated intelligent search engine
View/ Open
Date
2022-02Author
Arikosi, Michael Okurut
Karungi, Lydia
Mwanje, Mike
Kabahinda, Rovence
Metadata
Show full item recordAbstract
The knowledge of the law is not only important for the protection and enforcement of rights, but also for people to have some knowledge of the human rights obligations imposed on them by law. The purpose of this project was to develop and simplify the method of retrieval of up-to-date constitutional information. Furthermore, this project aimed at bridging the gap between the citizens in need of legal services and the qualified legal practitioners offering them.
In Uganda, the constitution information is not widely available and very few Ugandans know what is in the Constitution. Access to affordable legal services is another prevalent problem in Uganda. This is mainly because of the communication gap between the citizens and the legal practitioners.
With the advent of the internet, a solution to this problem was the creation of an easier means of information retrieval which is also capable of providing easy access to qualified legal practitioners.
Constitutional Awareness Dedicated Intelligent Search Engine (CADISE), is the solution we developed to solve this problem. It enables easy querying of the constitutional information as and when needed by the user. This was greatly achieved by the use of spaCy’s NLP model which is a free, open-source library for advanced Natural Language Processing (NLP) in Python. This model was used to tokenize the main constitutional text resource and extract sentences from the resultant document based on the user’s query string.
The ranking of the search results was done through a BERT (Bidirectional Encoder Representations from Transformers) sentence model. This model determined the similarity of the query string and the extracted sentences and ordered the results starting with the most similar to the least.
CADISE also allows citizens to contact qualified legal representatives registered to offer pro bono services on the system via email. A user can send an email to a registered legal representative, he or she can also make an appointment. The appointment functionality was achieved through the integration of a third party free online appointment scheduling software called Calendly.
In conclusion, the system was projected to be a time-saving data retrieval tool for legal information for academic and professional use. It should also be easily scaled up to a variety of information sources.