Image Mining on Facebook. Building a Webscraper using Python

      Аннотация: This work aims to attain two research goals. Firstly, a literature review on image mining applications is conducted primarily in the context of management and business and secondarily in other areas. The review makes clear that literature about image mining applications in the area of management and business is scarce. Secondly, an existing Python code is extended to enable it to scrape images and videos at large scale from Facebook and analyse them with Google Cloud Vision. Several challenges appeared during the development process such as the occurrence of error messages or the Application Programming Interfaces (API) dependency of the Python web scraper. Nevertheless, the code serves its purpose and is capable of downloading and analysing the images and videos of the Facebook posts from the provided input files. Moreover, the results can be uploaded to a Mongo Database (MonogDB). The author comes to the conclusion that the Python code can improve existing studies and with slight adjustments could even make infinite research areas accessible.

Ключевые слова: Image Mining, Facebook, Webscraper, Python