Content Analytics and Event Analytics are two different approaches to analyzing and detecting spam calls in a communication network. But how do they different, and when should you use which type?
In this article, we cover:
Content Service Providers (CSPs) are required by law — and urged by their clients — to mitigate spam and fraudulent phone calls. But this can be a complicated, expensive effort even for the largest CSPs, and sometimes it’s nearly impossible for the smaller CSPs to handle without a specific, enterprise-specific partner organization.
In any spam and fraud prevention tactics for CSPs, analytics are required. Generally speaking, there are two main models, which can be used independently or in combination with each other: Content Analytics and Event Analytics.
Content Analytics is a method of analyzing the content of a call, such as the audio content for a voicemail, or the text of an SMS message, in order to identify and flag it as spam.
The Communication Service Provider or it’s technology parter analyze the content of the call to determine if it is likely to be spam. Content Analytics uses sophisticated computer processing - such as natural language processing (NLP) and machine learning algorithms — to analyze the call content and make a determination as to whether it’s legitimate or likely spam or fraud.
This information can then be used to block a call, redirect it to voicemail, or label it (potential spam, spam, fraud, telemarketer, etc.)
Furthermore, some Content Analytics providers use AI-based algorithms to create a digital fingerprint of audio recordings, which means that bad actors — who rotate through phone numbers quickly — can be identified quickly, regardless of the number or the network they are using.
For example, a Communication Service Provider (CSP) may use a Content Analytics vendor, such as YouMailPS, to detect spam calls. YouMailPS gathers billions of data points on calls via it’s sister community company, YouMail, which offers voicemail & spam-blocking services to consumers. With this data, their AI/ML Content Analytics algorithms analyze the content of the call. If the caller is making a sales pitch or attempting to solicit money, the NLP algorithms used in Content Analytics can detect these patterns and flag the call as spam or fraud. Depending upon the product, you may receive a fraud risk score, a fraud campaign descriptor, or it can trigger automatic actions.
Likewise, a CSP or its vendor may also have fingerprints of authentic calls from a brand, and may compare calls that mention the brand to the fingerprints of the brand’s actual calls, to better rank some calls as fraud.
This “ground truth” is extremely valuable, claims Gerry Christensen, the Vice President of Enterprise Identity Defense at YouMailPS. He continues, “In fact, the USTelecom Industry Traceback Group relies upon YouMail content for objective evidence to initiate actions to stop bad actors and their campaigns.”
Event Analytics is a method of analyzing the events and data surrounding a call, such as the caller's number, the time of the call, the length of the call, the number of calls from that number, the location of the caller, and any other call patterns in order to identify and flag it as legitimate, spam, or fraud.
Event Analytics uses techniques such as statistical analysis, machine learning algorithms, and big data analytics to identify patterns and correlations in the data that are associated with spam calls.
For example, a CSP may use Event Analytics to label a certain number as likely-spam because (a) it has been ranked as “spam” by a higher percentage of users than normal, and (b) the call lengths are very short, and (c) there is an extremely high volume of calls.
However, this type of analytics does not get to the actual intent of the call. What if, for example, the number in question was actually a pharmacy, with automated calls letting their customers know that their prescription is ready. Perhaps the outgoing message is unclear, and for the first few seconds it appears as if it’s spam, which is why it may be tagged by some users.
Event analytics, while incredibly useful, does not provide as much information as a Content Service Provider should have to decide what to do with every call.
How They Differ And When You Should Use Content Analytics Versus Event Analytics
While both Content Analytics and Event Analytics are methods used by CSPs to detect and prevent spam calls, they differ in the approach they take and the types of information they analyze.
Content Analytics focuses on analyzing the contents of a call to determine if it is likely to be spam, while Event Analytics focuses on analyzing the events and data surrounding the call to determine if it is likely to be spam.
Content Analytics is better at determining intent, whereas Event Analytics simply describes what has been happening.
But the question as to “which should you use, Content Analytics or Event Analytics” is a bit of a false choice.
The best combination of services is clearly to use both Content Analytics and Event Analytics in combination with each other.
A number which is flagged by Event Analytics with high-risk factors can then easily be paired with Content Analytics data to label it as fraud — where each type of analytics, alone, may not have come to a conclusion with the same level of accuracy.
By using a combination of both Content Analytics and Event Analytics, CSPs can more effectively detect and prevent spam calls, ensuring that their customers receive only the calls that are important to them … and protecting CSPs from complaints, legal action, and unwanted FCC attention.
YouMail Protective Services helps CSPs minimize tax scam calls, fighting fraud on their networks. Get a free whitepaper on The Critical Role of Audio Analytics in Robocall Threat Mitigation Programs here >