Every year Aptoide opens its doors to the eager and disruptive minds of Portugal’s top college students for the Summer Internships. The goal is for the student to work on a specific project for two months, and many times the results are either put into production or result in a new business! In fact, Aptoide itself started as a Summer Internship project of its mother company, CaixaMagica.
We took this opportunity to learn more about a specific summer internship project that is being developed by Sara Silva. The project is the creation of a new social graph analysis system for the Aptoide user stores and is being developed under the supervision of Diogo Pires and Luís Pinto. Diogo has contributed to the current version of Aptoide Sentinel and is now working on Aptoide's Big Data platform. Luís is one of the creators of the MarDev and Topidis system for the Aptoide Influencer Program along with Claudia Fernandes.
We've gathered the three of them to answer some of our questions. Read on!
1. Hi Diogo, Luis & Sara! Please tell us a bit about yourselves - what do you do at Aptoide?
Sara Silva - Hi there! I'm Sara and I’m here as a Summer Intern. I have a degree in Computer Science and I’m currently finishing my Master’s in Information Systems and Knowledge Management. This is my first professional experience and I’ m working on a project related to graph network visualization for the Aptoide apps timeline.
Diogo Pires - Hi! I'm a Python Developer in the Backend Team. Our team handles several different systems running in the background, such as the automated malware detection system codenamed Aptoide Sentinel. Apart from that we also maintain the Big Data Platform that gathers, processes and stores all the data produced by our users. This data is then used for internal reporting through the Business Intelligence (BI) platform, the Push Notification system and to leverage the user experience in the app. At the moment I am responsible for the background systems and the implementation and maintenance of the Big Data platform.
Luís Pinto - Hello there! I'm a Marketer in the Community Marketing Team. I'm mainly responsible for the technological and analytical problems of the team. My background is in Marketing, IT and Statistics, and I fill this hybrid role between the three things I guess. The Community Marketing team handles all the content on this blog, our social media channels, and two community engagement programs: the influencers and the ambassadors. For the past year, I've been mostly involved in the development of the influencer marketing system, which employs social network analysis to measure the relative importance of different YouTubers and to measure the performance of the campaigns we run over time.
2. Sara, you've joined us here at Aptoide as a Summer Intern to work on a very special project with a huge potential for Aptoide. Can you please explain what it is all about?
Sara Silva - This project consists of developing a system to visualize graph networks. These graphs are made up of nodes, that represent the Aptoide stores that are assigned to a specific user, and by edges, links between stores. When two stores are linked, it means that one user is following the other or that they're following each other. To each edge we will assign a weight that defines how strong is the link based on how many things users have in common. We can understand and quantify relationships between users by mapping and analyzing these graphs.
3. What has been your biggest technological challenge so far and how did you overcome it?
Luís Pinto - So far we haven't yet encountered any major technological challenges. But I guess, once the system is in production, scale could become an issue. The amount of data we have available can make our social networks rise to CERN level problems really fast!
4. Can you share with us how do you think this graph network analysis applied to recommender systems will change how we work at Aptoide?
Sara Silva - With recommender systems, we will be able to improve user experience while they interact with Aptoide. The user will be able to visualize who is following him and who he follows. Also, knowing to which community the user belongs to allows us to make app recommendations based on the preferences of their friends. This recommender system helps us to satisfy the user needs by giving them a richer experience.
Diogo Pires - This project is very interesting and can produce important results for the recommender engines. For starters, it can make us understand better how and why our users are connected to each other. Compared to other Android app stores, Aptoide has a unique component in the form of a social network. Since Aptoide is all about apps, we can use our social network to helps us solve one of the biggest problems app stores have: App Discovery. Analyzing the social network can help us understand why the users are connected to each other by analyzing the installed apps and the categories of those apps, just to begin with. It can help us detect new features to use in the recommender engines and enable us to provide them with recommendations of connections, which we don’t currently have. In addition, it can help us identify why users aren’t connected but that a priori we think should be, thus identifying possible improvements in usage flows inside Aptoide.
Luís Pinto - Apart from the recommendations I also think this system can also be employed to measure the relative influence of our users. This will allow us to identify the opinion leaders of communities with similar app interests. This can be interesting to help us reward and attract interesting creators and power users, similarly to what already happens on other social media platforms! At the end of the day, people don't just want automated recommendations, they want to hear what these influencers are saying about the apps, so we need more of them to enrich our community.
5. What exactly do you hope to achieve by the end of these two months here at Aptoide?
Sara Silva - After these two months, I hope to make a productive contribution that favors Aptoide towards getting to know the profile of their users and improving their overall experience. This internship has proven quite valuable so far, contributing towards my personal objectives and to a great growth in my professional career. I truly hope that this project will impact Aptoide in a great and positive way.