Natural language processing, with its potential to connect non-computer-oriented users with the data they generate, is becoming a highly sought-after technology, not least by Facebook and Google as they race to harness the potential of artificial intelligence.
By definition, natural language processing (NLP)Opens a new window is the ability of a computer to assess both the meaning and intent of human-generated text.
â€œNatural language processing helps computers read and respond by simulating the human ability to understand the everyday language that people use to communicate,â€ says Marco VaroneOpens a new window , chief technology officer for Expert Systems, an NLP company. â€œWithout natural language processing, artificial intelligence only can understand the meaning of language and answer simple questions, but it is not able to understand the meaning of words in context.â€
More than half of a group of companies that describe themselves as aggressive cognitive technologies adopters say they are already using natural language generation and processing in their analytic platforms, according to a 2017 Deloitte survey.
Semantics Are The Rainbow’s End
The appeal of natural language generation is that it processes the data and information generated by a user through speech, and then informs the user of the potentially valuable information and insights that can be gleaned from this speech.
â€œInstead of users telling the software what they are looking to find, autonomous capabilities serve up insights based on identified correlations and patterns,â€ the MIT Technology ReviewOpens a new window reported in a study on machine learning-driven analytics. â€œThe result will be simplified and more personalized insights that anticipate requirements and make recommendations using predictive analytics.â€
The resulting analytics can help inform decision-making and carry out tasks like planning or reaching a conclusion based on only partial information, areas previously considered to be out of reach due to the requisite â€˜human’ intelligence.
Companies are increasingly exploring the broad range of potential business applications for natural language processing, including tailored customer service and gathering market intelligence.
Facebook has experimented with the potential commercial applications for natural language processing. These capabilities were the underlying technology for its â€œM suggestionsâ€ service, which has been integrated into Facebook Messenger. Based on the text surveyed in Messenger, M Suggestions offers fully automated suggestions for payments and making plans.
Facebook had originally envisioned developing a more comprehensive text-based virtual assistant, but the beta version did not become as popular, or commercially viable, as envisioned.
â€œIt’s possible to imagine a world where M was more successful at commerce and was able to take a cut of revenue, defraying some of the costs of maintaining an around-the-clock service,â€ says Casey NewtonOpens a new window , a writer for Verge who was given access to an earlier version of M that offered various shopping suggestions among its services. â€œBut bot-based commerce has been slow to take off, as most people continue to prefer native apps and the web over sending text messages.â€
Another company, Fin.com, is also using the natural language processing algorithms to sell its personal assistant potential more aggressively for office use, offering a range of services for $150 a month, from workflow customization to calendar integration and information-sharing across teams.
Better Understanding Of The Doctor’s Notes
The technology is also helping decipher doctors’ notes, using the resulting data to provide valuable research information on a range of diseases, including breast and other cancers.
â€œA key challenge to mining electronic health records for mammography research is the preponderance of unstructured narrative text, which strikingly limits usable output,â€ says Tejal PatelOpens a new window in an article for the American Cancer Society. â€œIn the era of EHR systems, big data, and machine learning algorithms, natural language processing has emerged as a possible solution with which to overcome the limitations of manual data abstraction.â€
Natural language processing cannot yet replicate the human ability to integrate and adapt to new information, known as natural language understanding, which will be the next big hurdle for language algorithms.
â€œMost of the methods employed in NLP are statistical in nature, and statistics can only go so far without context or semantics,â€ says Marco LagiOpens a new window , an MIT researcher on machine learning in Emerj. â€œThe algorithms behind the applications described above simulate human understanding and can do that at scale, but they are still brittle in that they can’t simulate a behavior they haven’t seen before.â€