>
Project: Rtdb

The Reptile Database

BA Thesis - from mockups to implementation. The Reptile database provides a catalogue of all living reptile species and their classification. Work included implementation from scratch utilizing Elasticsearch for full-text and spatial queries in addition to species range maps.

About

The Reptile Database(Wikipedia) Existing since 1996 this database provides a catalogue of all living reptile species and their classification with the goal of providing access to data on biodiversity. To extend the use of the database I redesigned the application from scratch and improved the user interface significantly. Thereby I was able to achieve good results developing a species-identification tool as well as a geospatial search.

Redesign

In the new version all previous features have been retained and new features have been introduced.

ID-Tool

To simplify the acess to catalog data a mechanism was needed to search the database without known scientific species names. Therefore, we created an identification tool based on morphological features, such as color, size, pattern and geo-location.

In close contact with the Reptile Database research team I came to the conclusion that an AI-based approach won’t be possible to the many natural deviations in reptile photos. I was able to utilize a search engine specialized on full-text search to rank results based on distinct morphological features achieving an identification-rate of 73% which is 10% less error-prone than todays AI-based identification approaches.

Range maps

Visualized spatial data

Derived from a research paper on the distribution of reptiles among the planet I published in Nature ecology & evolution from 2017 I could store the spatial data of ~9500 species. Consisting of vectorized coordinates they are used for location approximation in geolocation-searches.

I really enjoyed the visually appealing presentation of scientific research results, which are otherwise only available in scientific papers.

Implementation

As ~30% of the users browse the website on mobile devices it’s fully mobile compatible.

Mockups were created with the helped of knowledge gained from research questions from previous studies and use-cases from my thesis. Additionally I conducted a survey asking for the most desired features.

Not only are images crawled from other publicly available databases, catalog data now also features range maps for nearly every species and “RedList” tags for endangered species.

Font

One of the first time that I could gain knowledge by using specially selected typography in "the real world", with * Raleway* being the font of choice. Raleway delivers a professional look while keeping text clear and readable making it a good fit in an academic context.

Color

Main colors were abbreviated from natural elements and used in a combination of cool, rational tones of grey and mostly white to emphasize an academic background.

Motivation

The first real-world project were I was not only part of development but also intensely enganged in UI design and the visual appearance of an organisation.

Coming from over 70.000 lines of CSV files and two gigabyte of raw spatial data I am fascinated by this data transformation to achieve great visual output by having an eye on usability.

Old website