merterm/etymon

Find the origin of words in every language using a Deep Neural Network trained to create an etymological map.

Etymøn

Etymøn is an analysis and tracing tool for word origins in all languages. It will be used to review current etymological language families and if possible find new connections that were not already present in current taxonomy. It will accomplish this using a deep learning approach.

Etymøn will be a system that contains three main components: deep-learning, augmented-reality with object recognition, and generative dreaming. Etymøn bases most of its functionalities on the etymological map that will be generated by the deep learning algorithm. Etymøn system will include a palpable product in the end as an online application based on this etymological map. The system and its implementation will be broken down into three stages in respect to its components. Check out the website at http://etymon.org

Data Use

We have scraped data from several etymology websites. You can find them helpful in your projects. You can find the scraped data in csv format in the data folder. You can also find the whole dataset that Etymøn uses in the releases.

Group members

Nashiha Ahmed
Mert İnan
Cholpon Mambetova
Utku Uçkun

Supervisor

Professor Mehmet Koyutürk

Jury Members

Professor Uğur Doğrusöz
Professor Varol Akman/ Associate Professor Çiğdem Gündüz Demir

Innovation Expert

Dr. Çağla Çığ Karaman

etymon's Issues

LSTM through node.js

Currently Etymon's LSTM uses Keras and tensorflow. On localhost, they run on python virtual environment. python-virtualenv package is used for node.js to create a python virtual environment on server yet it does not work properly. Hence, we need a way to run LSTM on the server

etym.org doesn't work: backend is offline

Describe the bug
search hangs forever on http://etymon.org/website/etymon/index.html

console says:
2VM61:1 POST https://etymon-190009.appspot.com/search 404

it looks like your backend is offline