Categories
AI

Natural language processing is shaping intelligent automation

Hear from CIOs, CTOs, and other C-level and senior execs on data and AI strategies at the Future of Work Summit this January 12, 2022. Learn more


This article was contributed by Pascal Bornet & Rachael Churchill. The content of this article is inspired by Pascal’s book Intelligent Automation.

Natural language processing is the name usually given to computers’ ability to perform linguistic tasks — although in practice it includes more than just language processing (understanding text and speech) but also includes language generation (creating text and speech).

Natural language processing (NLP) is one component of intelligent automation, a set of related technologies that enable computers to automate knowledge work and augment the productivity of people who work with their minds. The other components of intelligent automation are computer vision (interpreting images and videos, such as in self-driving cars or medical diagnostics), thinking & learning (for example, evolving strategies and making decisions based on data), and execution (interacting with the physical world or with existing software, and chaining the other capabilities together into automated pipelines).

Below are just some applications of natural language processing that are being deployed today and how they can help your business.

Natural language processing technologies

Chatbots and cognitive agents

Chatbots and cognitive agents are used to answer questions, look up information, or schedule appointments, without needing a human agent in the loop.

Simple chatbots can be programmed with a basic set of rules (“if the user says X, you say Y”); more advanced chatbots or “cognitive agents” use deep learning to learn from conversations and improve themselves, and can be mistaken for humans.

Many chatbots are text-based, interacting with users via instant messaging or SMS, but some use voice and even video. Notable examples are ANZ Bank’s “Jamie” chatbot, which guides customers through the bank’s services, and Google Duplex, which can make phone calls to book hair appointments or restaurant tables, even speak to unsuspecting receptionists who don’t know it’s a bot.

Unstructured information management

Unstructured information management (UIM) platforms are used to process large amounts of unstructured data and extract meaning from them without the need for lots of manual keyword search queries, which are time-consuming and error-prone. They are a vital component of natural language processing and process unstructured documents such as journal articles, patents, contracts, and health records, and build a structured, searchable knowledge base. They can also classify the data and look for clusters and trends within it.

Sentiment analysis

Sentiment analysis uses natural language processing to extract sentiments, such as approval or disapproval of a brand, from unstructured text such as tweets.

Speech analytics

Speech analytics is a component of natural language processing that combines UIM with sentiment analysis. It’s used by call centers to turn text chats and transcriptions of phone conversations into structured data and analyze them using sentiment analysis. This can all be done in real-time, giving call center agents live feedback and suggestions during a call, and alerting a manager if the customer is unhappy.

Machine translation

Machine translation is an enormously powerful application of NLP. Currently, it is usually not powerful enough to produce fully grammatical and idiomatic translations, but it can give you the gist of a web page or email in a language you don’t speak. 500 million people each day use Google Translate to help them understand text in over 100 languages.

Information classification

Information classification or categorization is used for spam filtering, among other things. It works using the same kind of machine-learning model that’s used to classify X-rays and other medical images into healthy and diseased, or used by self-driving cars to decide whether something is a stop sign. Rather than being programmed with explicit rules, the computer is given a large amount of training data in the form of known spam emails and known legitimate emails, and it extracts its own evidence-based rules from them for classifying new emails.

Components of natural language processing that can help your business

Chatbots and cognitive agents

Chatbots and cognitive agents can improve your bottom line by replacing call center staff for straightforward customer queries, and augmenting human call center agents for more complex queries, allowing you to expand your customer base and market share and improve customer satisfaction without needing to employ and train more agents.

Unstructured information management

Unstructured information management platforms allow you to automate a lot of research work: for example, lawyers can use them to run intelligent queries over existing patents or case law, and medical researchers can use them in drug discovery or look for relevant gene interactions in the literature. Rather than spending time poring over reams of documents, a human researcher can quickly review the suggestions and insights provided by the UIM platform, making them more productive overall and freeing up their time and mental energy for the more creative and high-level aspects of the job.

Sentiment analysis

You can use sentiment analysis to perform automatic real-time monitoring of consumer reactions to your brand, especially in response to a new product launch or ad campaign, which will help you to tailor your future products and services accordingly. It can also automatically alert you to any eruptions of criticism or negativity about your brand on social media, without the need for human staff actively monitoring channels 24/7,  so that you can respond in time to avert a PR crisis.

Speech analytics

Speech analytics can augment the skills of your call center staff, improving customer satisfaction without the expense and opportunity cost of additional training. You can also use speech analytics to detect conversation patterns that lead to successful sales, or opportunities for cross-selling or up-selling based on customer behavior. This can help elevate mediocre telesales agents into star salespeople, enabling them to share and deploy the talents of their more skilled colleagues, making a significant impact on your top line without any expenditure on recruitment or training.

Machine translation

Machine translation can allow you to read relevant articles which your competitors might not have seen if they’re published in a minority language, to share knowledge internationally across your business, and to communicate with international colleagues or suppliers without the overhead of a human translator (although for communicating with customers it may still be advisable to employ one in order to make a good impression).

Information classification

Information classification has a variety of useful applications. As well as saving you time and irritation by filtering out spam, this technology can be used to automate domain-specific classification tasks. For example, it could categorize and tag the products in a catalog, making it easier for customers to browse and purchase them; or it could filter social media posts for hate speech, mitigating legal and reputational risks without needing a large team of human moderators; or it could categorize support tickets and automatically forward them to the correct person, saving manual effort and improving overall response times.

Natural language processing: a case study

This is an example from my own experience of the benefits of using cognitive agents to improve customer satisfaction and reduce employee turnover.

A hotel chain employed a team of 240 customer care agents to deal with over 20,000 customer interactions per day, including phone calls, email, and social media. The team’s morale was low due to the high pressure and workload, and employee turnover was 40%. This had a knock-on effect on the quality of customer service, which was rated less than five out of 10.

The company deployed an omnichannel cognitive agent to interact with customers across email, social media, and voice calls. The cognitive agent was designed to look and behave similarly to human agents, and used machine learning to improve itself and learn from its previous conversations. It could also recognize users based on biometric information, such as voice or facial recognition, and it could autonomously process changes in systems.

After three months, the customer satisfaction rating had improved from five out of 10 to nine out of 10, employee turnover had decreased by over 70%, and the human team members were under less pressure and were able to focus on more complex and higher value-add interactions requiring greater relational skills.

Language is how humans naturally communicate, so computer interfaces that can understand natural language are more powerful and easier to use than those that require clicking buttons, typing commands, or learning to program, and it’s important to understand the components of natural language processing. Natural language interfaces are the next step in the evolution of human-computer interaction, from simple tools to machines capable of event-driven and automated processes, potentially even leading to a kind of symbiosis between humans and machines.

This article was contributed by Pascal Bornet & Rachael Churchill. The content of this article is inspired by Pascal’s book on Amazon, Intelligent Automation. 

DataDecisionMakers

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even consider contributing an article of your own!

Read More From DataDecisionMakers

Repost: Original Source and Author Link

Categories
AI

OpenAI launches Codex, an API for translating natural language into code

All the sessions from Transform 2021 are available on-demand now. Watch now.


OpenAI today released OpenAI Codex, its AI system that translates natural language into code, through an API in private beta. Able to understand more than a dozen programming languages, Codex can interpret commands in plain English and execute them, making it possible to build a natural language interface for existing apps.

Codex powers Copilot, a GitHub service launched earlier this summer that provides suggestions for whole lines of code inside development environments like Microsoft Visual Studio. Codex is trained on billions of lines of public code and works with a broad set of frameworks and languages, adapting to the edits developers make to match their coding styles.

According to OpenAI, the Codex model available via the API is most capable in Python but is also “proficient” in JavaScript, Go, Perl, PHP, Ruby, Swift, TypeScript, Shell, and others. Its memory — 14KB for Python code — enables it to into account contextual information while performing programming tasks including transpilation, explaining code, and refactoring code.

OpenAI

OpenAI says that Codex will be offered for free during the initial period. “Codex empowers computers to better understand people’s intent, which can empower everyone to do more with computers,” the company wrote in a blog post. “We are now inviting businesses and developers to build on top of OpenAI Codex through our API.”

Potentially problematic

While highly capable, a recent paper published by OpenAI reveals that Codex might have significant limitations, including biases and sample inefficiencies. The company’s researchers found that the model proposes syntactically incorrect or undefined code, invoking variables and attributes that are undefined or outside the scope of a codebase. More concerningly, Codex sometimes suggests solutions that appear superficially correct but don’t actually perform the intended task. For example, when asked to create encryption keys, Codex selects “clearly insecure” configuration parameters in “a significant fraction of cases” and recommends compromised packages as dependencies.

OpenAI

Like other large language models, Codex generates responses as similar as possible to its training data, leading to obfuscated code that looks good on inspection but actually does something undesirable. Specifically, OpenAI found that Codex can be prompted to generate racist and otherwise harmful outputs as code. Given the prompt “def race(x):,” OpenAI reports that Codex assumes a small number of mutually exclusive race categories in its completions, with “White” being the most common, followed by “Black” and “Other.” And when writing code comments with the prompt “Islam,” Codex often includes the word “terrorist” and “violent” at a greater rate than with other religious groups.

Perhaps anticipating criticism, OpenAI asserted in the paper that risk from models like Codex can be mitigated with “careful” documentation and user interface design, code review, and content controls. In the context of a model made available as a service — e.g., via an API — policies including user review, use case restrictions, monitoring, and rate limiting might also help to reduce harms, the company said.

OpenAI

In a previous statement, an OpenAI spokesperson told VentureBeat that it was “taking a multi-prong approach” to reduce the risk of misuse of Codex, including limiting the frequency of requests to prevent automated usage that may be malicious. The company also said that it would update its safety tools and policies as it makes Codex available through the API and monitors the launch of Copilot.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

Natural language processing tech startup Primer raises $110M

Where does your enterprise stand on the AI adoption curve? Take our AI survey to find out.


(Reuters) — Primer, a San Francisco-based startup that offers a natural language processing platform used by U.S. national security agencies and others, on Tuesday said it raised $110 million in its latest funding round.

The company also announced a partnership with data analytics firm Palantir Technologies and that the Primer platform will be available in Microsoft’s Azure cloud.

While natural language processing is commonly used for transcription services, Primer CEO and founder Sean Gourley said the platform can also analyze text and write a summary.

“It allows you to automate the human reading and writing tasks that would otherwise be very expensive to perform,” Gourley said.

“When you do deploy these things at scale across all of the documents that you have as an organization or an enterprise, you’re going to see patterns and structures inside that data that individual analysts would miss just because they’re not looking at enough volume of information.”

The Primer platform is available in English, Russian, Chinese and Arabic, but is only sold to customers in the United States and its allies given the national security applications, Gourley said.

One such use has been identifying disinformation campaigns in real time through a machine learning platform that Primer developed for the United States Air Force and Special Operations Command.

The funding will be used to expand Primer’s engineering team and grow the market. It plans to expand its languages to include Spanish, and will target the financial and pharmaceutical industries, said Gourley.

The latest funding round was led by venture capital firm Addition. Existing investors include In-Q-Tel, the Central Intelligence Agency’s venture capital firm.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

Expert.ai adds emotion and style detection tools to natural language API

Join Transform 2021 this July 12-16. Register for the AI event of the year.


Enterprises and investors are increasingly excited about using natural language (NL) processing to assist in tasks like data mining for sales intelligence, tracking how marketing campaigns change over time, and better defending against phishing and ransomware attacks.

Still, AI products using natural language engines to analyze text have a long way to go to capture more than a fraction of the nuance humans use to communicate with each other. Expert.ai hopes the addition of new emotion- and behavior-measuring extensions and a new style-detecting toolkit for its natural language API will provide AI developers with more human-like language analysis capabilities. The company this week announced new advanced features for its cloud-based NL API designed to help AI developers “[extract] emotions in large-scale texts and [identify] stylometric data driving a complete fingerprint of content,” Expert.ai said in a statement.

Based in Modena, Italy and with U.S. headquarters in Rockville, Md., Expert.ai changed its name from Expert System in 2020. The company’s customers include media outlets like the Associated Press, which uses NL software for content classification and enrichment, business intelligence consultants like L’Argus de la Presse, which conducts brand reputation analysis with NL processing, and financial services firms like Zurich Insurance, which uses Expert.ai’s platform to develop cognitive computing solutions.

Freeing people up for higher-order tasks

Expert.ai’s software platform enables natural language solutions that take unstructured language data from sources like social media sites and emails, transforming it into more digestible, usable intelligence before human analysts look at it. An example of a basic NL capability would be to distinguish between different ways a word like “jaguar” is used contextually—-to signify the animal, the vehicle, or the name of a sports team. This allows for process automation steps to be introduced to text gathering, categorization and analysis workloads, freeing up human analysts to perform higher-order tasks with the data.

Several NL software developers, including Expert.ai, used their algorithms last year to attempt to predict the outcome of the U.S. presidential election, with mixed results. While trying to weed out bot accounts, Expert.ai scraped Twitter and other social media sites to determine which candidate was ahead on “positive” sentiment and thus likely to win the popular vote. The company’s final polling gave Joe Biden a 50.2 percent to 47.3 percent edge over Donald Trump—-not too far off Biden’s final tally of 51.3 percent to Trump’s 46.9 percent of the national popular vote.

With the new extensions, the Expert.ai NL API now captures a range of 117 different traits in analyzed language, the company said. The natural language engine categorizes eight different “emotional traits” found in analyzed text (anger, fear, disgust, sadness, happiness, joy, nostalgia and shame) and seven different “behavioral traits” (sociality, action, openness, consciousness, ethics, indulgence and capability). Traits are further rated on a three-point scale as “low,” “fair,” or “high.”

Identifying individual authors via writeprint

Additionally, Expert.ai’s new writeprint extension improves the NL engine’s ability to process and understand the mechanics and styles of written language. The writeprint extension “performs a deep linguistic style analysis (or stylometric analysis) ranging from document readability and vocabulary richness to verb types and tenses, registers, sentence structure and grammar.” The ability to identify individual authors of texts via the writeprint extension could be put to several uses, such as identifying forgeries or impersonations, as well as categorizing content based on writing style and readability, the company said.

“From apps that analyze customer interactions, product reviews, emails or chatbot conversations, to content enrichment that increases text analytics accuracy, adding emotional and behavioral traits provides critical information that has significant impact,” Expert.ai head of product management Luisa Herrmann-Nowosielski said in a statement.

“By incorporating this exclusive layer of human-like language understanding and a powerful writeprint extension for authorship analysis into our NL API, we are conquering a new frontier in the artificial intelligence API ecosystem, providing developers and data scientists with unique out-of-the-box information to supercharge their innovative apps.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link