Creating a proper data management strategy for a company is not easy. There are many types of databases and approaches you may choose from. We would like to tell you more about graph database use cases and business problems that you may solve by leveraging this type of database in your organization. Read more.
What is a graph database?
Graph databases allow you to safely store data without losing any complexity. Instead of using rows and columns, this type of database stores data as relationships between entities, and represents information in the form of nodes and edges. Such an approach to storing data can be compared to making a map of some ideas on a blackboard, using keywords and arrows to express the relations. In graph databases, your data doesn’t have to be changed and limited to the information that fits a predefined model.
Graph database use cases in banking
Fraud of many types is a great concern of the modern financial market. To improve fraud detection, graphs can be created in order to visualize relationships between entities Once such a graph is ready, a quite simple query can be run in order to find similarities in users accounts, such as e-mail addresses, passwords, personal details, etc. With this method, you can find accounts held by the same users and even detect money laundering.
It works similarly when detecting money mules, people who transfer illegal money from their own account to a different scam operator (who often resides in another country). Using machine learning for fraud detection is becoming more and more popular, but continuing to develop better models is not easy as user accounts usually provide limited information. Graph technology enables us to get more data from each account because we are able to visualize and assess the relationships between accounts and transactions. It allows us to learn which accounts are most likely to be mule accounts. Using graph technology, you can significantly enhance your ML-based fraud detection system.
Banks and other financial institutions also create other opportunities for fraud themselves. They all try to provide their customers with the best user experience (UX), and they often try to do so by simplifying the verification process during transactions. Giving up some additional steps of verification makes it easier for criminals to access the transaction, copy it and fraudulently show a fake transaction to the customer. Graph technology enables fast answers to queries and as graphs contain a lot of data, this solution quickly became popular in the world of real-time fraud detection.
Graph database healthcare use cases
Graph networks are also leveraged in healthcare, especially in precision medicine and experimental medicine, where there is a strong need to analyzing huge amounts of information. With graph technology, you can create many networks, for example:
- a drug network – in which each node stands for a drug, and there can be various relations between nodes,
- a disease network – which contains information on many diseases and the connections between them,
- a gene network – which stores information about genes and how they are related,
- a patient network – which can consist of patients’ family history, symptoms, or other important information.
All these networks can be brought together as a dataset full of details about entities and relations which could never have been created with different technology and which can be used to answer even the most difficult medical questions. Algorithms can be applied on graphs and networks to reveal useful insights.
Graph technology can also be leveraged for drug retargeting/repositioning. So-called orphan diseases are certain, rare conditions that affect only a small group of people. As these people constitute a rather small market for new meds, pharmaceutical companies do not always prioritize developing such medications. In such situations, attempts are sometimes made to use a pre-existing drug to treat the disease. Analysis performed on graph networks can help identifying opportunities to use a medicine to treat a different disease than the one for which it was originally designed.
With graphs, patterns can be found faster and with higher efficiency. The combination of graph technology, advanced algorithms and artificial intelligence can improve the capability of designing treatments, but also to predicting the possibility of getting sick or having a genetic disease.
Graph databases in e-commerce
Graphs provide your company with a more complete view of your customers’ data. Businesses gather huge amounts of varied data every day – such as:
- personal data (like name, age, gender, address etc.),
- transactional data (purchases, time of finalizing the transaction, types of items etc.),
- behavioral data (related to how a user behaves on the Internet).
These are just a few examples, of course. There are many types of data that can be used to better understand your customers. Just as in the case of using graph databases in healthcare, various networks can be created to analyze complex relations between nodes.
You can use your knowledge about various groups of your customers to define segments and personalize content, but also to provide your e-commerce users with product recommendations, which can improve not only your sales, but also user experience. Of course, there are many types of non-graph technologies that could support recommendation engines. Still, graph databases, and the algorithms you can apply to the data stored in them, allow your recommendation systems to produce more accurate suggestions faster.
Creating product recommendations in real time requires the ability to correlate different types of information (for example user data, product details and behavioral information) in a very short time. A graph database is the right solution to bring all this data together and form connections in no time because you can store the data in a form that contains the information about the relationships between entities.
Graph databases for social networks
Social media sort of became a miniature of the real world. It can be used for communication, entertainment, selling things, declaring your political opinions, working or educating yourselves. Imagine all those connections between users, images, videos, groups, pages and other elements of this virtual reality. There are several challenges that can be overcome by leveraging graph databases in social media.
“Sockpuppet” accounts… In addition to stealing accounts and data, there are also fake accounts run by bots. What for? Actually, there are many reasons. Likes, comments, shares and other actions affect content’s popularity on social media. Owners of fake accounts use bots to generate fake reactions to posts and other published content in order to create trends, increase sales or achieve any other goal. In some countries, such activity can be even leveraged to cause social unrest, destabilize governments and share hate or prejudices.
Graph databases are used by many of the most popular social media platforms to gather information about the relations between users. This way, they are able to learn which accounts may be run by bots and check them. Various relations between users and published content are also leveraged for content recommendations.
Are you wondering if a graph database may be a good solution for your business? Contact us. We will be happy to analyze your business needs and advise you on the best technologies for your company.
Check out our blog for more details on Big Data:
- What is Big Query, and how can it support your analytics?
- Why is BigQuery replacing Hadoop for enterprise analytics?
- Optimizing Apache Spark