Helicopter View: Natural Language Processing applications are conquering the enterprise. Slowly
A New Normal is emerging as NLP finally allows direct interaction between computers and human beings through the use of voice or text commands. From how we work to how we order our groceries, this changes everything. Or at least, it will.
NLP tools have evolved rapidly over the few past years, as machine learning algorithms open up a whole new world of possibilities for applications which had previously been based on strict “if…this…then…that” (IFTTT) manually written rules. These rules were unable to reach the wide range of real-life language possibilities but that’s all history with the advent of machine learning tools.
Today, professionals working with NLP combine artificial intelligence tools and computational linguistics to develop ways in which machines can understand inputs given in human languages and perform tasks or verbal responses as a response to these.
At the most basic level, an app either places a delivery order or tells you where the local pizzaiolo is working his magic.
However, the near-horizon promises so much more than this. In addition to its applications in consumer gadgets, (such as home, voice, and smartphone assistants like Google Home, Amazon Alexa, Siri, et al), NLP is also finding its place within enterprise settings, with integrations in tools that can facilitate life within the workplace.
NLP is becoming a useful integration for enterprise tools ranging from email auto-correction and spam filtering to language translation and sentient analysis.
Let’s take a quick look at some examples of how NLP is already performing in the wild;
IBM: Financial NLP applications
IBM is using NLP tools to provide a vast amount of services for financial enterprises.
These include services such as IBM’s AlchemyData News API which financial firms rely upon to write a query that will create alerts when analysts upgrade or downgrade stock. Similarly, News API can send instant alerts about crucial changes to a company’s management.
IBM’s Alchemy DataNews API also allows financial companies to monitor industry events where they have exposure or material interest; oil spills would be one such instance. And then there’s AlchemyLanguage (through Watson Developer Cloud) which aids the sector in monitoring sentiment related to earnings or acquisitions, tracking any positive or negative reactions to these.
Similar NLP tools also allow financial institutions and banks to track general customer sentiment about their services by analyzing and monitoring social media posts.
MindMeld: NLP for voice recognition
MindMeld develops tools which client companies deploy to create their own voice-driven interfaces for apps and devices. In doing so, the company uses NLP technologies that can understand human language and answer users’ questions.
MindMeld has received widespread recognition as a leader in natural language understanding, and was named by Entrepreneur Magazine as one of 100 Brilliant Companies in 2015.
Auto-correction and Text Prediction
Another application of NLP tools for enterprise products and services is text prediction; the process that allows for text auto-correction and prediction of users’ most used words or emoticons.
SwiftKey is just one example of a business that develops text-prediction technology using NLP. They have a team of some 150 people (and one inflatable dinosaur), operating out of their Central London HQ and bases in San Francisco and Seoul. However, the space is white-hot at the moment so expect news of plenty more startups muscling in on the Sector.
NLP for Social Media Analysis
NLP can also be useful for companies wishing to analyze Social Media and gain insights into the public’s general opinions about them.
IBM is not the only company providing tools for SoMe monitoring and analysis. Others, such as NetBase, also provide similar tools which, in the round, are not only extremely valuable for qualitative market research purposes, but are also making their presence felt in political campaigning.
In fact, Cambridge Analytica (CA), the data firm credited with helping Donald Trump win the US election (Paywall), is also seen by many observers to have played a decisive role in the final days of the 2016 Brexit campaign where Britain voted to leave the European Union.
Question Answering and virtual assistants
Just as voice-operated assistants such as Siri and her pals are rapidly being integrated into consumers’ homes, cars and devices, similar tools seem set fair to make their way into the enterprise space.
Question Answering (QA), provided by voice assistants or chatbots, could help employees manage their daily tasks and use their time more efficiently.
Unified Communications service, Slack, recently integrated bots such as ‘To-do bot’, ‘Statsbot’, and ‘Growbot’ to help teams manage their workload, access statistics, and generally achieve more in less time.
NLP can help make such tools more responsive and efficient by effectively interacting with their human users to support them throughout their working day.
The future of NLP
These are only a few examples of applications of NLP within the enterprise space, and the variety of solutions that use these tools to assist businesses in their work is likely to grow further over the next few years.
In a blog post, IBM references recent report by MarketsandMarkets predicting that the NLP market will reach $13.4 billion by 2020, at a compound annual growth rate of a stratospheric 18.4%.
However, developers with NLP skills still have to face challenges related to language processing, context understanding and developing machines that use traceable reasoning strategies. And, those developers are still frustratingly thin on the ground; we discuss that in more depth here.
There can be no doubt that we are well on our way into the era of the NLP-Normal but these skills shortages are impacting directly on the pace of change.
Indeed, reaching critical mass is at the heart of the question. And, it is worth a helluva lot more than $64,000.