Natural Language Processing (NLP) may seem the least notable of Semantic technologies and their applications, particularly when considering the hype surrounding graph databases, Cognitive Computing, and the Internet of Things. Nonetheless, Markets and Markets reports that the NLP market is projected to grow to $13.4 billion–with a CAGR of 18.4 percent—in the next five years.
Technological mainstays such as Google, IBM, Microsoft, and others are all making lucrative investments in this space, so it might be difficult to understand why NLP doesn’t garner the sort of headlines that other Data Management technologies and applications routinely do. The answer is simple: it is so deeply ingrained in some of the more salient aspects of Data Management that the substantial value it adds to them oftentimes goes unattributed.
Natural Language Processing and its spawn, text analytics, play invaluable roles in:
- Social media sentiment analysis: NLP is able to distinguish positive comments from negative ones, even in instances when slang is used and when positive terms are used negatively (and vice versa).
- Business Intelligence: According to Forrester, contemporary BI tools incorporate user interfaces that exploit NLP to offer a more self-service experience.
- Data Discovery: Search and all of its myriad Internet-based applications are greatly attributed to NLP and text analytics.
- Data Governance: The combination of NLP and text analytics is frequently used to detect instances of aberrations in regulatory compliance, especially in heavily regulated industries such as finance and health care.
- Cognitive Computing/Artificial Intelligence: NLP enables machines to transmit data to language and vice versa.
An examination of these and other ways in which NLP enhances data-driven processes for the enterprise further substantiates its role as one of the more important aspects of Data Management in contemporary times.
The Relationship Between NLP and Text Analytics
Although these terms are occasionally used interchangeably, text analytics is a subset of NLP and is one of the two forms of analytics options NLP provides. The other is speech analytics, which many consider a separate capability altogether for the simple fact that the former is widely considered to be still developing. NLP is responsible for determining relationships in documents, performing search, pinpointing and understanding boundaries of sentences and phrases, and determining names and places via Semantic technologies. When applied to text analytics, NLP can facilitate uses cases such as identifying aspects of regulatory compliance, performing sentiment analysis, categorization, and text clustering. Text analytics can extract specific words from sources that pertain to specified objectives. Most NLP solutions are either statistical based, rule based, or a hybrid of the two.
Social Media Sentiment Analysis
Other than general search, the most frequently used application for NLP is for sentiment analysis of social media. Such analytics can inform Business Intelligence processes and clarify meaning from all the various slang terms and multiplicity of references to a particular product or service that the enterprise offers. NLP can identify the sentiment of these feeds (and other sources of text) for different languages while generating reports, answering queries (particularly via Cognitive Computing applications), and extracting desired information. One of its primary advantages when used with social media is its ability to effectively create structure from unstructured, external Big Data, which in turn can be combined with proprietary on-premise data to provide comprehensive overviews of customers, products or services. Text analytics of social media sentiment is one of the most readily accessible wins for NLP when utilized with Big Data. This application is valuable across vertical industries but particularly informs aspects of marketing and advertising.
Pattern and Image Recognition
Some of the more cutting edge applications of NLP involve its correlation with image and pattern recognition. This aspect of NLP is related to another application of Semantics, Machine Learning, which in turn involves aspects of Cognitive Computing and Artificial Intelligence. While text analytics and speech analytics are separate capabilities of NLP, this underlying processing can also be applied to derive words from images and, at some point in the future, perhaps even create images from words. Utilizing various aspects of Neural Networks and Deep Learning, NLP was recently leveraged by researchers in Silicon Valley to apply descriptions to images. Previously the image recognition of NLP was limited to simple objects; recent research has indicated that it can provide descriptions of details in those objects, such as the type of clothes a man is wearing, colors, and comments about the background. Although this aspect of NLP is still being developed, the implications are that it can determine even more meaning from unstructured data in the form of videos and still images—to be combined with the enterprise’s structured data. The recognition capabilities of NLP also include optical character recognition and interactive voice response
Cognitive Computing and Artificial Intelligence
The uses of NLP in Cognitive Computing are myriad. In addition to discerning patterns among images and relating them to language, NLP enables users to submit codeless queries in natural language that are answered that way. Certain Cognitive Computing platforms (such as IBM’s Watson Analytics) and other analytics options provide narrative driven explanations of queries to substantiate findings. When combined with traditional dashboards and visualization tools, these capabilities can help to further the art of Data Storytelling and lessen the gap between analytical insight and action. By providing the means by which machines can communicate with humans on conventional language-based terms, NLP’s position in Cognitive Computing is invaluable. Markets and Markets indicates that: “NLP is considered as a sub-field of artificial intelligence and has significant overlap with the field of computational linguistics.”
Life Sciences and Health Care
Some of the more recent headlines related to NLP pertain to its many applications in the fields of health care and life sciences. NLP has many uses in these fields, including facilitating more expedient discovery of drugs and their uses, monitoring quality of various forms of treatment, and extracting important textual information from numerous research documents and reports for the purpose of data mining. Several of these uses actually involve sentiment analysis in which organizations are monitoring social media streams for key words and phrases denoting the efficacy of follow up procedures, treatments, and the presence of disease outbreaks. Forrester revealed that when combined with various facets of Cognitive Computing and Artificial Intelligence, NLP can foster:
“Capabilities that can assist and offer independence to the sight-, hearing-, or mobility- impaired, including helping those with visual impairments recognize text and products and allowing the hearing-impaired to listen and speak via unconventional media such as bone conduction earpieces.”
Data Science and Beyond
Whether used in conjunction with Cognitive Computing or not, most NLP applications involve some form of Machine Learning or conventional algorithms that require the specific tailoring of Data Scientists. These professionals can tailor NLP for log analysis of security models, risk management and regulatory compliance, and price and demand forecasting—in addition to the typical sentiment analysis and search capabilities for which it is known. The reality of NLP is that it is an integral component in most aspects of Data Management today. Its propensity for quickly determining relationships in text and extracting crucial information substantially adds to everything from predictive analytics to Data Governance while enabling numerous facets of Big Data, Cognitive Computing, and Data Science. Its refinement can help to extend its capabilities to more sophisticated forms of image recognition and speech capabilities, which can very well revolutionize the analytics process as it is known today.
