Life as an intern @ BearingPoint NL

In this blog I will describe my experience as an intern at BearingPoint NL! This blog will focus on:

  • the goal of this research internship;
  • my daily activities as a research intern;
  • how this research may benefit the client;
  • reasons for doing an internship at BearingPoint.

My name is Luca De Vita and I did a six-months internship from February 2020 until July 2020 at the Data & Analytics (D&A) team. This internship was for my master’s graduation project; the goal of this internship was to combine academic research with real-world problems. 

An introduction into natural language processing

Businesses are increasingly interested in using natural language processing (NLP) techniques to extract relevant information from the plethora of textual data. However, many businesses struggle to get insights from textual data because this data in inherently unstructured, so traditional data analysis techniques cannot be used. For this reason, businesses contact the D&A team of BearingPoint to help them with these challenging endeavors.

The goal of my research

My graduate research project was focused on finding the natural language processing (NLP) techniques that performed best for a client in long-term healthcare. Afterwards, my colleagues could use this information to improve an existing knowledge graph for the client.

An important subfield in NLP is called keyphrase extraction. The goal of keyphrase extraction is to develop algorithms that extract the most “relevant” phrases in a document. Keyphrases can concisely summarize the context of a document, but keyphrases have a wide variety of purposes. More specifically, keyphrases are often used as metadata in document management systems or knowledge graphs. In short, this allows the employees of Vilans to quickly retrieve relevant documents, which can increase the employee productivity. Additionally, it could empower faster and more informed decision making for healthcare professionals.

To successfully complete this project, I had to work extensively with programming languages like Python, but there was also plenty of opportunities to gain insight into other technologies like graph databases (Neo4J). Therefore, my daily activities consisted mostly of programming, but also other research related activities like reading scientific papers and writing my master’s thesis.

The results of my research were that combining several NLP techniques could give state-of-the-art results. Furthermore, the utilization of Neo4j also allowed me to tackle one of the main challenges in this research field: extracting keyphrases that are not present in the text, but effectively summarize the document.  By using this approach Vilans and healthcare professionals can quickly retrieve relevant documents, which leads to increased productivity.

There are four aspects I liked about doing my internship at BearingPoint. Firstly, there was a good balance between developing my hard and soft skills. For example, I learned much about NLP and programming, but could also develop my presentation skills. Namely, I could present my research within the firm and to the client. Secondly, there is good supervision from more experienced colleagues. My supervisor often helped with tackling challenging problems in my research, but also gave me feedback on my presentation skills. Thirdly, the D&A team consists of young and ambitious people, which I personally really liked. Furthermore, there are many social activities organized throughout the year, so you can also meet colleagues from different teams. Lastly, BearingPoint is an international company, which is located in 22 countries. Therefore, there is the possibility to work in an international setting, and potentially abroad.

To conclude, I can highly recommend doing an internship at BearingPoint. My internship was mostly focused on doing research, but BearingPoint also offers opportunities to work as a working student intern. Thus, there are many possibilities to do an internship at BearingPoint and develop yourself!