Protect against Fraud with Entity Extraction

Fraud is a significant source of financial and reputational loss for organizations. In fact, the Association of Certified Fraud Examiners (ACFE) estimates that a typical organization loses 5% of revenue every year due to fraud.

Fraud can take many forms including financial, health care, credit card, insurance, Medicare/Medicaid, billing, tax, lottery, etc.  Unfortunately, every type of organization is susceptible to fraud. And those organizations that are unprepared to defend against fraud are likely to experience proportionally larger revenue losses. Detecting fraudulent behavior is therefore critical to protect your organization’s finances and reputation.

How to Fight Fraud with Entity Extraction

To get an idea of how fraud can impact your business and how to defend against it, consider financial fraud.  Analysts who work for financial organizations battling financial fraud have to get on top of enormous volumes of data if they are going to uncover fraudulent actions, unearth hidden links between accounts, and track relationships between actors engaged in the fraudulent activity.  In addition, though much of the data is structured, a large and growing amount of relevant data is coming in the form of unstructured data.  Examples of such data include:

• Suspicious Activity Reports produced by financial institutions when they identity potentially criminal activity involving money transfers
• Phone call records or emails between customers and the financial institutions
• Any relevant information to be found on social media

Advanced text analytics technologies like Entity Extraction can expedite the processing and review of unstructured data and free up analysts to do more complex analytical work.  It can also facilitate the discovery and connection of critical information from seemingly unrelated data sets.

Why Use NetOwl’s Entity Extraction

There are a number of ways in which NetOwl offers unique Entity Extraction capabilities critical for financial fraud analysis:

1. NetOwl identifies an extensive ontology of named entities with state-of-the-art accuracy, including key concepts such as names of people, organizations, phone numbers, money amounts in all currencies, addresses, etc.
2. NetOwl offers a unique and advanced capability to identify a broad range of relationships (e.g., person-associate, person-affiliation) and events such as financial transfers as well as the participants like who made the transfer, which institutions were involved, and the dates on which they occurred.  Once these concepts have been identified, the information becomes available not only to search but can also be discovered through advanced analytical tools such as semantic indexing, link analysis, and social network analysis.  These relationship and event extraction capabilities allow for a far more advanced analysis beyond the simple links afforded by co-occurrence. They enable a deeper analysis of documents that makes possible the application of a technique such as network link analysis that will reveal clues critical to a given financial fraud investigation. The ultimate goal is the identification of suspects and their associates as well as patterns of fraudulent behavior.
3. NetOwl offers name normalization so that names can be more easily resolved and aggregated across documents.
4. NetOwl is engineered specifically for high-volume processing of multiple different data sources. Additionally, NetOwl integrates document converters to handle hundreds of native document formats.
5. NetOwl integrates easily with databases, document and content management systems, portals, and other sources of electronic content.  Its REST API supports easy integration into existing workflows.  NetOwl can be deployed on premise or in the cloud for horizontal scalability, offering rapid processing of massive amounts of data.

In summary, NetOwl offers an extensive set of Entity Extraction capabilities to turn unstructured data into actionable insights that speed up fraud investigations and make them successful.