Go Beyond the Tip Of The Iceberg with Semantic Analytics
Today, with the ongoing explosion in the volume of data collected, many organizations are living in irony when it comes to translating all the data into meaningful actions. The irony is gathering too much data but knowing too little about what this data represents. To address this emerging issue, many analytics platforms powered by cognitive computing technologies have started going beyond conventional text analysis to deep semantic analytics to provide an in-depth understanding of, say, emotions within languages and a customer’s response to them within a marketing scenario. Therefore, the premise of this article is to explain how businesses can detect new opportunities and the gaps in the markets they serve more quickly and at a lower cost by tapping into semantic analytics.
Semantic Analytics Explained
The main actor behind the scenes of semantic analytics is meta-knowledge which includes the origins and nature of the data, and deep knowledge about how to portray it. There are chances, the sounds of meta-knowledge reminded you of metadata but meta-knowledge provides you with the relationships between a data element and other data, context about its meaning, while metadata provides basic information about data, such as who created it, its format and a brief description of its contents. What I mean by that is, say, you are searching for all customers who play string instruments, then the platform will demonstrate a list that includes people who play the violin, viola, cello, bass, harp, and dulcimer, but not drums or trumpet. One of the reasons why this relationship of a data element to other, related data elements would be valuable for a business may be to identify prospects with similar interests or characteristics for creating personas for a specific marketing campaign.
On a more relatable note, when you make a query about a keyword on Google, as you may have noticed, the platform performs the search results based on not only that keyword but also commonly used synonyms, abbreviations or even misspellings of frequently used terms for the word. This performance also represents one of many valuable use cases of semantic analytics.
How Semantic-Layers Work
You may ask how that information is being funneled into a user-friendly interface. The vehicle is semantic-layers. The main purpose of semantic-layers is to translate technical jargon into recognizable business names and organize them into a format that seems logical to the front-line business person. In Wikipedia’s words:
“A semantic layer is a business representation of corporate data that helps end users access data autonomously using common business terms. A semantic layer maps complex data into familiar business terms such as product, customer, or revenue to offer a unified, consolidated view of data across the organization.”
The concept of the semantic layer was originally patented in 1991 by Business Objects. However, of course, since then the volume of data generated has been in an explosion. In fact, IDC estimates that the amount of data generated by everything from point-of-sale systems to sensors will rise from 4.4 ZB in 2013 to 44 ZB in 2020. Therefore, a semantic model that once was originally mostly static, has become dynamic to handle a growing number of data. Besides the purpose of scaling, as advances in data science are sparking more creative business opportunities, semantic analytics platforms have started illustrating the breadth and depth of excitement about artificial intelligence and deep learning in recent years.
To put semantic layers into perspective, we can look at how a semantic layer of loan data would make it easier for ad hoc questions to be asked by the lending team, or how a marketing team can take advantage of a semantic layer that focuses on campaign management or geo-demographic analysis.
How Semantic Analytics Deliver Compelling Business Outcomes
Even though there are numerous areas where organizations could take advantage of semantic analytics, in this article, we will specifically look into business outcomes from marketing, enterprise search, and security perspectives.
From a marketing perspective, social media metrics such as the numbers of clicks, mentions, retweets, shares and such are not sufficient anymore to capture the full potential of consumer data. Social media sentiment, on the other hand, is a great way to complement the information about how often your audience talks about your brand and understand the context behind those mentions. “Social media sentiment is the perceived positive or negative mood being portrayed in a social media post or engagement,” says Nick Martin, Social Engagement Coordinator at Hootsuite.
The primary purpose here is to understand the emotion behind a post, and it is critical as it enables brands to avoid sharing any tone-deaf social media posts in an event of a large outcry in response to the product. Or even better, a brand can predict this situation before it happens by taking advantage of sentiment analysis during their market research prior to the launch. Proactiveness that sentiment analysis provides may also be provided when marketers are trying to come up with the right message for a marketing campaign. It is always beneficial to take the temperature of their target audience before sending out any message.
Semantic Analytics runs multiple concept maps to analyze the same data using different terms, or the use of the same terms in different contexts. As a result, marketers can infer whether their use of a term means their audience is pleased or displeased. Capturing subtle differences in the language used provides a competitive edge specifically for marketing. With that capability, organizations could find what works most effectively to inspire a positive action. For companies leveraging direct response channels such as retail, eCommerce, telecom, travel/hospitality, media/entertainment, and financial services, adopting sentimental analysis tools can give them an upper-hand in the market that they serve.
It’s worth noting that sentiment analysis based on social media is only one aspect of the whole concept. Depending on the needs of a business, it may be wise to go beyond social media sentiment as organizations can miss out on fully unleashing the potential of data as it is often limited to binary choices, such as positive vs. negative.
According to a recent study by IDC, “The High Cost of Not Finding Information,” the average knowledge worker spends up to 2.5 hours per day searching for or gathering information or data. This includes searches, email queries and other related tasks that all result in a massive amount of time spent trying to find information that already exists. This equates to approximately 400 or so hours per employee, per year searching or gathering information.
The study also shows if a worker cannot find the information they are seeking within 4 minutes they will either recreate it, use older content assets or interrupt a co-worker. It costs companies, not only because of the extended time workers spend searching for information online, which is estimated at least 10 hours a week, but also because of the fact that between 3% and 5% of an organization’s files are lost or misplaced. Annual losses to a Fortune 1000 company with one million files is $5M, according to Information Week data. On the other hand, if the median Fortune 1000 company were to increase the usability of its data by 10%, company revenue would be expected to increase by $2.02 billion, based on InsightSquared’s study.
The transformation from traditional keyword-based enterprise search engines to semantic search has started off with injecting Natural Language Processing (NLP) into the enterprise search platforms. From there, other AI technologies such as machine learning and natural human interactions have waded into the foray in time. Today, thanks to the advent of new technologies, cognitive search platforms are capable of predicting the intent behind the search request and increasing the relevancy. As a result, cognitive platforms now are enabling the identification and surfacing of intelligent content in context to any business application able to consume it.
With nearly 1 million new malware threats released each day, detecting security threats in complex IT environments is not an easy task. However, it is arguably the most important business-critical task of a modern business for obvious reasons. Security is an issue for organizations that concerns not only internal but also external shareholders. In fact, 51 percent of Global Citizens with internet access are discouraged by the lack of trust online, choosing to not engage in online transactions due to privacy and security concerns, a new Ipsos survey for the Centre for International Governance Innovation has found.
Semantic analytics activates automated systems to go beyond a simplistic check of whether, for example, traffic to or from a given port falls outside a normal range. Additionally, it enables such a system to learn which combinations of dozens of network characteristics are most likely to indicate an attack, and which other metrics it should check if one measure falls outside the normal range. The best part is that as the automated systems learn about new types of threats, or gain more insights into older threats, semantic analytics makes it easy to add new systems, behaviors or threat types to the analytic process.
The Next Market Wave
According to Gartner, as the visual-based exploration paradigm has become mainstream, a new innovation wave is emerging that has the potential to be as disruptive as (or more than) visual-based data discovery has been to the previous semantic-layer-based development approach of traditional BI and analytics platforms. The Stamford, Conn.-based research firm sees Natural-language processing (NLP), natural-language query (NLQ) and natural language generation (NLG) for text- and voice-based interaction and narration of the most statistically important findings in the user context as key capabilities of smart data discovery. However, the tasks for creating insights are still largely manual and prone to bias, according to Gartner’s analysts, and the solution to address this issue is, going forward, accelerating data harmonization and the visual identification of patterns in data through visual-based data discovery approach instead of the previous IT-centric semantic-layer-based approach.
Before exploring what the next smart data discovery market wave is, in my opinion, organizations, first, should get the basics right. There is nothing more impulsive than seeing a promotion about the company when you are experiencing an issue with them but it still happens all the time as tone-deaf social media posts are all over the place. Organizations should adopt technologies that can identify the sentiment and context within unstructured information and use it to improve their organization. To me, plain content is not king anymore as long as it is not turned into a self-describing resource. Thanks to the advent of technology, today’s cognitive platforms’ natural language processing is capable of assigning topics against a pool of 25 million semantic entities, by weight. So why not gain insight at a fraction of the cost and time without the restrictions of traditional methods and taking the guesswork out of business processes?