toern

Project result: Poster (PDF)

Quality assessment of fashion returns in e-commerce using machine learning techniques

Description of the company
toern has made it its mission to rethink return management and make it more sustainable. With our solution, we enable online shops to forward suitable returns directly to the next customer. With this process we are able to save large amounts of CO2 emissions and costs for the online shops.

Situation
To forward suitable returns directly to the next customer, we need to ensure the quality of returns from the original buyer.
As soon as a return is not of the desired quality, it makes more sense to send it back to the online shop's return center in order to save on unnecessary detours.
It also prevents customers from having a negative experience with our service.

Currently only the reasons for the return, provides information about the condition. So the idea is to let customers take an additional picture of the returns.
We are currently focusing on fashion returns.

Problem
The problem now is to determine the quality of the returns based on the picture.
The output should be used as an additional factor to check whether an item is suitable for onward shipment.

Aims of the project
Build a machine learning model to assess the quality of fashion returns in e-commerce based on a set of
features.

• Collecting a set of fashion return images (e.g. by web scraping)
• Pre-process of the collected data and labeling the dataset according to their quality
• Evaluate the model’s performance using different metrics (e.g. accuracy, precision, recall)
• (Optional) Analyze the model’s predictions to identify patterns or insights about the features that contribute to a good or bad quality return.

Scopes
Hands-on experience in data preparation, feature extraction, machine learning modeling and evaluation.

Target group (students)
Students who are interested in data science, machine learning and e-commerce.

Study programs: computer science, data science or related fields.

Additionally students who are interested in the intersection of technology and business, particularly in the
e-commerce industry.

Skills: First experiences with python (or another programming language), data science and machine learning.
Additionally students who are interested in using machine learning to solve business problems.

Dates
Please save these dates: Fishing for Experience Termine

Registration
You can apply for Fishing for Experience online.