- In Archive
- 2017-10-20 22:34:48
- 510 Views
[caption id="attachment_37156" align="aligncenter" width="1024"] Machine learning presentation; Courtesy of Petar Petrov[/caption]
In the summer of 2017 two club members of The Hub, Martin Nestorov and Dimitar Kumanov, joined the Documaster team as part of an internship. On Oct. 12 they presented the projects and shared the experience that derived from their work at the organization. Nestorov and Kumanov elaborated on the topics of machine learning and natural language processing[NLP]—a field of computer science, artificial intelligence, and computational linguistics concerned with the interactions between computers and human languages.
The whole matter developed from the two students taking a one credit course under Dimitar Ouzounov, Head of Development at the Documaster Application Server, who at that time was looking for potential candidates for the internship at Documaster. The requirement for the internship was the creation and evaluation of a small project. Both Nestorov and Kumanov met the criteria and got accepted as the only interns at the organization. There, they were appointed to develop a stand-alone project, a task they had no experience with. Nevertheless, they successfully managed to make a natural language processing app over the course of three months with the guidance of the Documaster team.
Nestorov and Kumanov stated that being part of the Documaster team has created opportunities for them in terms of gaining specific skills that could only be acquired through working in a real-world environment.
"After we dove into the meat of things, we got to learn about so many different technologies and techniques - from working with Git(control system designed to handle projects) in a company environment to constant experimentation with different kinds of NLP libraries, everything was extremely beneficial for us”, said Nestorov.[caption id="attachment_37155" align="aligncenter" width="1024"] Students at the Machine learning presentation; Courtesy of Petar Petrov[/caption] In their presentation in front of the students, Nestorov and Kumanov demonstrated the operations performed by the application they developed during their internship. The application evaluates articles from popular news sites and lets the reader know what the articles are about with the use of keywords. These keywords are found through a specific set of algorithms they created. Users can then further type a keyword and find related articles. Based on the user's reading habits, the application provides fresh recommendations that are compatible with the user’s preferences. If the user finds the recommendations unsatisfactory, the application can get feedback through its Graphical User Interface (GUI) and process it.
"Advanced work with Git, setting a whole database, data mining and cleaning, and developing algorithms for phrase extraction, are just a few of the things we got from this whole project," commented Nestorov.