Different analytics platforms will have different natural language features and implementations. It’s important to understand the various natural language terms so that you can determine the right analytics solution for your needs.
Natural Language Processing
Natural language processing (NLP) speaks to a machine’s ability to understand human words or speech. Historically, computers have understood coding languages and mathematics. If someone wanted to build a webpage or run a program, they’d need to know HTML, Clojure, or one of the myriad other coding or programming languages. Now, with NLP, computers are reading our language.
In an analytics platform, NLP enables a user to ask a question just as you would in a conversation. For example:
- What were sales last year by brand and category?
- What is penetration for Brand A by state?
In these instances, NLP allows users to easily ask day-to-day questions and get the information they’d normally receive in a routine report.
Advanced NLP platforms can go even further, answering more complex questions like:
- How did Brand A perform last month?
- Why did performance decrease?
The “how” and “why” questions are more challenging and require more advanced machine learning to answer.
NLP platforms that can understand questions that are less cut-and-dry allow business people to spend more time acting against the information they learn and setting strategy accordingly.
NLP is not to be confused with keyword-based search. Keyword-based search recognizes human words, but not in a natural, complete sentence. Think: “Sales, 2018, Brand, Category.” While keyword search is similar to natural language, it still requires a level of translation from the user, who must dice the flow of their thoughts into keywords the machine can understand.
Natural Language Query
A natural language query (NLQ) is the question that a person asks in normal human speech. NLQs do not require any special formatting, syntax, or vernacular.
In analytics platforms, NLQs refer to the questions that users are inputting into the interface.
Traditionally, business intelligence tools enabled a data scientist, data analyst, or technical user to pull dashboards together based on business data. With the rise of AI in analytics platforms, more business intelligence tools encourage NLQs.
This change has enabled non-technical users to ask the questions that come to mind (as opposed to thinking mathematically, selecting equations and special calculations to generate the right dashboards).
Natural Language Generation
Natural language generation (NLG) refers to a machine’s ability to output information in human speech. This means that a computer is speaking your language.
In analytics platforms, NLG usually refers to generation of insights. If you ask a question like “what were sales last month,” the machine should relay an answer like “sales in January totaled $1 million.”
NLG allows analytics tools to offer responses that are easy and immediately understandable. NLG tools that also leverage machine learning can even provide complex analysis, where a question may require paragraphs of in-depth information to adequately answer the query. This depth is invaluable because it provides a truly comprehensive overview of your data.
The combination of NLG and machine learning is referred to as augmented analytics.
NLG vs NLP: Infographic
To recap, NLG and NLP are both different, but connected concepts.
In the realm of business intelligence tools, back-end processing of languages like SQL allow these platforms to produce insights. From a user perspective, the experience is seamless. Users simply type questions in natural language and receive answers in the same fashion.
For more helpful images like this infographic, check out this free eBook, “The State of AI in Analytics & Business Intelligence“
Natural Language and AnswerRocket
AnswerRocket knows the value of natural language. That’s why our platform offers both NLP and NLG. And of course, we support and encourage NLQs.
Natural language facilitates faster workflows, organic problem-solving, and focus. When your team can put brainpower into questions and answers — instead of the translation process — they can spend more time acting on the information they learn.
After all, business people have skill sets for selling, marketing, setting strategy, and reaching customers. When important questions can be automated with natural language, business people can spend more time focusing on the skills they were hired for, instead of waiting for routine reports from a technical team member.
When we get inspired, we want to act immediately.
With AnswerRocket’s natural language capabilities, your team can quickly ask a question (from the simple “what”s to the complicated “why”s and “how”s) and get an answer in seconds. Their curiosity can take them farther as they ask follow-ups, read in-depth insights, and immerse themselves in your company’s data narrative.
