During pandemic period in Turkey, analysis the complaints to the cargo companies by machine learning methods

This study attempts to discover how cargo, transportation, and logistics firms, which have a considerable market share in Turkey's service sector, manage their processes during the Covid-19 outbreak and suggests solutions to the adverse outcomes. The data used in the research include complaints made for firms in the cargo sector from an online complaint management website (sikayetvar.com) from the beginning of the pandemic to the date of the research (11.03.2020 – 30.09.2021), which contains the words related to the pandemic were collected with Python language and Scrapy module web scraping methods. The complaints obtained were evaluated with qualitative methods, and the six most observed topics were identified. These topics are as follows; delayed or not delivered, not at home or not delivered", phones not answered, return process, not received or delivered, and hygiene rules not followed. Complaints were classified by using multilabel classification algorithms according to the analyzes obtained from the training data created according to the determined topics. According to the findings, it was observed that the most complaints were experienced during the delivery process of parcels, and the number of delayed issues was the highest.