Description (english)
The goal of the project is to develop a prototype for the administration and association of school datasets of the non-profit projects „Mathe im Advent“ of the company Mathe im Leben gemeinnützige GmbH and „Abiturpreis“ of the German Mathematicians' Association (Deutscher Mathematiker-Vereinigung e. V.).
Both projects have a large number of school records from different sources. Teachers also adjust and add new school records each year. Many school datasets for the same school with minor differences in name and address are created several times. A high coverage of schools simplifies usage of the two projects. Although both projects use similar data, there is currently no common interface for exchanging data.
To solve this problems a software prototype for the automatic recognition of school records belonging to the same school, their association and correction will be developed.
The core components are:
AI-supported or pattern-based search of similar school records (finding duplicate entries from the same school)
A form to merge similar school records into a school record
A form for separating definitively different school data sets
CSV import of school records from „Mathe im Advent“ and „Abiturpreis“
CSV export of current school records in the required formats for „Mathe im Advent“ and „Abiturpreis“
The participants structure and execute the software project according to Scrum. Product owners will be the main developers of the non-profit project „Mathe im Advent“ of the company Mathe im Leben gemeinnützige GmbH.
Depending on the interest and number of participants, the following additional features are implemented:
Reinforced learning for the AI-supported search for similar school records, based on the information provided in the association/separation forms.
Verification and correction of address data of school records e. g. via Google Map API
Different treatment of already checked and newly imported school records according to source
Creation of diffs and related histories and the possibility to undo associations
Web crawler to supplement the database from available website data
Subsequent further development and maintenance to a productively applicable software is possible and will be remunerated by the company Mathe im Leben gemeinnützige GmbH on a royalty basis.
Target group:
Bachelor's and Master's students of computer science or mathematics with a bachelor's degree in computer science and knowledge of Artificial Intelligence, Pattern Recognition or User Interface Design.