NLP, short for Natural Language Processing, may be 2023's biggest trend in analytics.
NLP is not new to analysis. A decade ago, ThoughtSpot built its entire platform around the concept of natural language search, allowing its users to ask questions about their data using words instead of code. Since then, most analytics providers have added at least some NLP features.
For example, Tableau launched Ask Data, Explain Data in 2019 and subsequently acquired data storytelling provider Narrative Science, leading tothe launch of Data Stories in 2022. Oh, the white thememade data storytelling a key part of its platformalso. And AWS, Oracle, and Qlik, among others, have strong NLP capabilities.
But in the next 12 months, NLP, which for many years has been a means of enabling some self-service BI but has grown slowly due to language complexities, could become the dominant trend in analytics, according to industry experts.
The reason: ChatGPT,a new chatbot that was released by OpenAIin November 2022, which dramatically advances the Q&A capabilities of chatbot technology.
“By 2023, natural language processing will be a breakthrough,” said Donald Farmer, founder and director of TreeHive Strategy. "Big language model tools like ChatGPT are becoming increasingly popular for gaining insights from unstructured data."
NLP is not expected to be the only trend in 2023. Data fabric and decision intelligence are also growing trends that could become more significant in the next 12 months.
Advancements in NLP
NLP holds great promise for analytics.
Ideally, a business user could type or speak a query using any sentence in any language, and your BI platform could respond by providing relevant data in a digestible format, such as a chart or graph, accompanied by a detailed explanation. In addition, the platform would be able to answer more detailed follow-up questions in the same way, so that the business user could drill down into their organization's data.
ButNLP has not achieved this ideal state..
The language (e.g. different ways of asking the same question and different words with the same meaning and words spelled the same way with different meanings) is still too complex for existing technology that must take natural language, translate it into code to run a query, and then translate a natural language encoded response for your consumption.
Also, there are more than 5,000 languages around the world. Even programming a platform to chat in some of the most widely spoken languages is time consuming and complex.
But technology is improving. And it could result in significant advances in the use and capabilities of NLP by 2023.
"Natural language generation," said Ritesh Ramesh, COO of healthcare consulting firm MDAudit and client of analytics provider ThoughtSpot, when asked what the top analytics trend will be in 2023. "The ability to automatically generate business insights and feedback behind the ideas [will be in fashion]. It's the technology that automatically generates the narrative from the visualization."
In particular, the launch of ChatGPT, based onthe capabilities of GPT-3which was first released in 2020, could drive NLP's widespread improvement in analytics.
Some providers have developed their own NLP technology. For example, ThoughtSpot had NLP capabilities when it first launched in 2015,A AWS criou o QuickSight Qwith Amazon's machine learning technology, and Qlik added NLP capabilities with the acquisition of CrunchBot in 2019.
Now others have the potential to more easily add NLP capabilities by integrating ChatGPT or another chatbot tool.
Like GPT-3, ChatGPT translates the written or spoken word into code, executes a requested query, and translates the response back into natural language. But ChatGPT does this in a way that represents a technological leap rather than an incremental improvement.
"We're seeing tremendous innovation in querying and natural language generation; leaps are being made," said Dan Sommer, senior director and global market intelligence leader at Qlik. "ChatGPT is extremely powerful and shows that this natural language chatbot interface will move the dial in the short term. It changes the way we interact with data and how data is generated for us."
Amazon, Microsoft and Tableau's parent company Salesforce are all expected to make substantial improvements to their NLP capabilities over the next 12 months, according to Farmer.
“These [new chatbots] can generate SQL from natural language input, making them more advanced than the previous generation of natural language query technology,” he said. "This is something that will likely be at the forefront of development in the coming year."
However, there is still a long way to go before NLP reaches its ideal state, according to David Menninger, an analyst at Ventana Research.
Like other NLP technologies including ChatGPThe doesn't really understand the language. It's still a computer program and can only do what it's programmed to do. This does not include interpreting the meaning when a query is not clearly stated.
Furthermore, the answers are not always accurate, and like other AIs,is subject to prejudice.
"NLP still has a long way to go before it becomes popular. But I think [the evolution] will continue to happen until 2023," Menninger said. "There are still many gaps."
While 2023 is not the year that NLP will completely transform BI, NLP technology will improve over the next 12 months, he continued.
"The things that I think will happen with NLP will be things like more fluent questions rather than structured questions, and I think multilingual will become more and more prevalent in 2023," Menniger said.
In addition, he highlighted that the advances in NLP that he expects for this year will befocus on chatbots-- the written word -- rather than the spoken word.
"I still think the voice is going to be delayed," he said. "It will continue to be more chatbot-oriented, but we'll see greater progression on the voice front."
While NLP will be a dominant trend in analytics in the coming year, it won't be the only one.
One thatrose to fame in 2022and should continue to gain momentum in 2023 is decision intelligence.
Decision intelligence is the use of augmented analytics and machine learning to help humans make business decisions. With the amount of data organizations collectgrowing at an exponential rate, there is too much data for even a team of data experts to sift through and notice every change or anomaly.
AI and ML, however, can be trained to monitor key metrics and other data for changes, anomalies, or trends; bring these perceptions to light; and alert data experts who can later determine whether action is required.
Vendors that focus on decision intelligence include Pyramid Analytics, Sisu Data and Tellius, all of which have attracted venture capital funding in the past 16 months despiteworsening conditions in the capital markets.
Pyramid raised $120 million in venture capital funding in May 2022, Tellius raised $16 million in October 2022, and Sisu last raised $62 million in late 2021.
"There are many opportunities to help organizations make smarter decisions," Menninger said. "I think we're at the beginning of the trend, but... companies don't get enough support today from vendors and technology in general."
This plays into what Krishna Roy, an analyst at 451 Research, callsactionable intelligence, which he said he expects to rank high in business analytics in 2023.
While not specifically decision intelligence,actionable intelligenceIt is related to decision intelligence in that it is the use of technology to support decisions that leads to action. But rather than displaying information for data specialists, the concept is to present information in business users' workflows.
"Actionable intelligence is going to be a big trend," Roy said. "Data-driven decision making will need to be done using multiple approaches depending on the use case and the user. Actionable analytics, as the name suggests, will enable this by allowing people to more easily perform the analysis by providing it into .. .familiar workflows they use regularly.
Qlik is one of the providersmake actionable intelligence a priority.
NLP and decision intelligence focuses on using technology to enhance human interaction with data.
Another trend that is gaining momentum has less to do with technology and more to do with a philosophical approach to analytics.
Data mesh, first introduced in 2019, is an analytical approach that decentralizes data. Although its benefits have been known for a few years, adoption has been slow. That could change in 2023 as organizational leaders look to new ways to make analytics a bigger part of their business.
Traditionally, organizations have stored data in centralized repositories overseen by a team of data experts, and the data is analyzed as needed. In some cases, even the analysis is handled by a centralized team that delivers on-demand reports and dashboards.
ANdata meshInstead, the approach allows teams in different domains, such as finance and marketing, to monitor and analyze their own data whileconnecting the data of each domain through a data catalog.
The idea is to leverage the domain knowledge of an organization's employees, operating under the assumption that a data specialist in finance or marketing will have more knowledge of financial or marketing data than a centralized data analyst.
And just like NLP and decision intelligence, data mesh also aims to expand the use of data to inform decision-making within organizations. It promotes the idea that domain experts can more easily work and teach self-service users how to work with data than a centralized team.
"We're going to see that the people who were talking about decentralization a year ago, two years ago, maybe even three years ago, were right," said Russell Christopher, director of product strategy atStarburst Data Mesh Specialist.
Christopher, whose experience includes six years at Tableau and 14 years at Microsoft, said he needed to show people the benefits of data visualization before they finally realized its value. Now, vendors like Starburst, Talend and Denodo, whose tools enable data merging, are doing the same for data decentralization.
"Now the light bulbs are starting to light up," he said.
Farmer noted that he is also seeing organizations exploring the data mesh, but added that it is still in its early stages. warned thatmore data governanceit must be included in the data mesh before it is viable at scale.
“In general, data mesh is a self-service type of BI, and some are exploring the architecture more rigorously,” he said. "But a major issue that still needs to be addressed is governance, which is likely to prevent companies from embracing the concept."