ABBYY, a Digital Intelligence company, today launched a global initiative to promote the development of trustworthy artificial intelligence (AI) technology. As AI becomes ubiquitous across consumer and enterprise high-value and large-scale uses and more open source tools become available for digitizing data, the ethical use of accessing and training data is imperative.

A growing number of technology leaders expect fair, transparent and ethical AI systems in order to fuel the continued adoption of AI and spur innovation. In fact, by 2025, Gartner estimates 30 percent of large enterprise and government contracts for the purchase of digital products and services that incorporate AI will require the use of explainable and ethical AI. Furthermore, three-fourths of consumers say they won’t buy from unethical companies, while 86% say they’re more loyal to ethical companies. Therefore, ABBYY made public its core guiding principles on developing, maintaining and promoting trustworthy AI technologies and advocates for other technology leaders to do the same.

“Innovation and ethics go hand in hand. As the use of AI proliferates, it is important for technology leaders to adhere to and promote the utilization of technologies that are transparent, fair, unbiased and respect data privacy,” commented Anthony Macciola, Chief Innovation Officer at ABBYY. “By adhering to high standards with regards to the performance, transparency and accuracy of our products, we are able to deliver solutions that have a tremendous impact for our customers.”

• ABBYY, whose Digital Intelligence solutions leverage AI technologies including machine learning (ML), natural language processing (NLP), neural networks, and optical character recognition (OCR) to transform data, affirmed its commitment to the following principles and advocates for other leading technology organizations to also commit to trustworthy AI standards:
• Incorporating a privacy-by-design principle as an integral part of its software development processes
• Protecting confidential customer and partner data
• Developing AI technologies that meet or exceed industry standards for performance, accuracy and security
• Empowering customers and partners to successfully implement digital transformation in their organizations by delivering solutions that provide a greater understanding of content and processes
• Providing visibility into the performance characteristics and metrics of its technologies, as well as providing opportunities for product feedback
• Delivering AI technologies that are socially and economically beneficial
• Fostering a culture that promotes the ethical use of AI and its social utility

“AI has the power to yield significant social and economic benefit,” continued Macciola. “With ethics in mind, we have the ability to transform the future in a manner that promotes innovation, accelerates technological advancements, and augments human intelligence, creativity, and capabilities responsibly.”

The post ABBYY Launches Global Initiative Promoting the Development of Trustworthy Artificial Intelligence appeared first on Artificial Intelligence.

The most appreciative advancements in statistical AI, the ones with the most meaning and potential to improve data’s worth to the enterprise, are deep learning deployments of computer vision and natural language technologies.

The distinctions between these applications involve much more than image recognition versus that of speech or language. Horizontal computer vision use cases pertain to some aspects of inter-machine intelligence, e.g., scanning videos or production settings for anomalies and generating alerts to initiate automated procedures to address them.

Conversely, natural language technologies provide the most effective cognitive computing application for furthering human intelligence, decision making, and the action required to extract business value from such perceptivity.

While the utility derived from image recognition largely varies according to the vertical, the capability for machines to understand natural language—for humans to interact with databases in layperson’s terms across sources—strikes at the core of converting the unstructured data of language into informed action.

Few organizations, regardless of their industry, could not benefit from this capacity. The application of deep neural networks and other machine learning models for this universal use case presents the greatest win for the enterprise, resolves the issue of unstructured data, and is currently taking the form of the following capabilities:

♦ Natural language generation: According to Forrester, natural language generation systems (such as those associated with Alexa and conversational AI systems) leverage “a set of rules, templates, and machine learning to generate language in an emergent, real-time fashion.” Accomplished solutions in this space rely on basic precepts of deep learning to generate text for an array of use cases.

♦ Smart process automation: The impactof equipping bots and other means of process automation with algorithms from cognitive statistical models is unprecedented. Instead of simply implementingthe various steps necessary for workflows, such approaches can actually complete them by rendering decisions conventionally relegated to humans.

♦ Spontaneous question-answering: Answering sophisticated, ad hoc questions across data sources has always posed a challenge for machine intelligence options. When backed by deep learning techniques and other aspects of AI, organizations can overcome this obstacle to profit from any unstructured, written data they have.

No one can deny the merits of deploying cognitive computing to accelerate data preparation or make back-end processes easier. However, the aforementioned applications of natural language technologies shift that ease and expedience to the front end. They’re the means of directly empowering business users with the peerless predictions of deep learning and, more importantly, redirecting its business value from fringe use cases to those impacting mission-critical business processes.

Natural language generation

When applied to natural language technologies, deep learning’s chief value proposition is the capacity to issue predictions— with striking accuracy, in some cases—about language’s composition, significance, and intention. Models involving deep neural networks facilitate these advantages with a speed and facility far surpassing conventional, labor-intensive methods of doing so. According to AX Semantics CTO Robert Weissgraeber, “Neural networks, trained with deep learning, are used in the generation process for grammar prediction, such as finding the plural of ‘feature’ or ‘woman.’”

Natural language generation has swiftly become one of the most useful facets of natural language technologies. Both Gartner and Forrester have recently developed market reports monitoring its progress. More importantly, it’s also revamping BI by accompanying everything from visualizations to reports with natural language explanations. Perhaps even more significantly, natural language generation-powered systems have expanded from conversational AI applications to include “product descriptions, automated personalized personalized messaging like personalized emails, listing pages like the Yellow Pages, and select journalism applications like election reporting, sports reporting, and weather,” Weissgraeber noted.

Natural language generation’s rise can be partly attributed to its extension of natural language processing (which is transitioning from being trained by rules to being to being trained by machine learning models) to include responses. Specifically, natural language generation employs natural language processing components such as dependency parsing and named entity extraction to analyze what the user writes, and then creates hints for the user to make his configuration faster, Weissgraeber explained.

The principal components of natural language generation systems include:

♦ Data extraction and generation: Thistool chain handles what Weissgraebertermed “classic data processing.”

♦ Topic- and domain-dependent configurations: Natural language generationsystems rely on this component to analyzedata’s meaning.

♦ Word/phrase configurations: Theseconfigurations are used to select differentphrases based on the desired meaning.

♦ Textual management: These elementsbring the “text together, with grammarprediction, correct sentences, text length,and formatting,” Weissgraeber said.

When combined with a system push or API for delivery, these natural language generation characteristics utilize deep learning for stunningly sophisticated use cases. Forrester indicates that in finance, this “technology can review data from multiple sources, including external market conditions and a client’s investment goals and risk profile, to produce a personalized narrative for each of an advisor’s clients.”

The post Next-generation natural language technologies: The deep learning agenda appeared first on Artificial Intelligence.

Data scientists are now part of the gig culture movement, but should you hire a freelancer instead of a full-time data scientist? If you lack data science talent or your existing data science team needs expertise it lacks, such as computer vision or natural language processing, perhaps you should consider a contract hire. But freelance data scientists aren’t always the answer to an organization’s needs.

Overall, companies are more inclined to hire a full-time data scientist than an individual contractor, if employment sites are any indication. On March 5, 2020, Indeed had 11,297 full-time data scientist positions listed and only 283 contract jobs. CareerBuilder had 3,122 full-time positions listed and 451 contract jobs.

Of course, there are other options. Employers could hire a part-time data scientist or use a consulting firm. If enterprises have a particularly compelling problem, such as curing a form of cancer, another option is to host a data science competition using a platform such as Kaggle.

However, before hiring data science talent, it’s best to understand what data scientists do because there are some nuances specific to this type of freelancer that hiring managers are wise to understand.

What is a data scientist?

A data scientist is a data expert who often holds an advanced degree in mathematics or statistics and probably knows how to code in R or Python. The most sought-after data scientists also have relevant business domain expertise.

While skill sets vary among individuals, a data scientist’s job is to help their employer solve difficult problems often involving discovery, optimization and/or prediction. The role may be considered part of IT or it may be specific to a departmental function. Of all possible data-related roles, data scientists tend to be the most sophisticated type of talent.

There are many myths surrounding data scientists, which can be counterproductive to hiring for the role.

The most common myth is the “unicorn” many organizations look for. This fictional character knows everything there is to know about data and is a coding superhero and a mathematical or statistical genius. Just point this individual at data and magic will happen.

This false belief results in unrealistic job requirements and unrealistic expectations of what data scientists and data science can do.

Why hire a freelance data scientist?

Matt Johnson, COO of data science consultancy Data Mettle, said there are three reasons clients tend to turn to freelance data scientists versus hiring full-time help: They aren’t sure they need a data scientist, they lack the expertise to understand what skills they need to hire or they just want to do a stand-alone project.

“Often, if they have some data and they think they can do something interesting or of value with it — rather than hiring a data scientist — it makes more sense to bring in someone for a few weeks or a month to explore the data, understand the business challenges and opportunities and what’s feasible,” Johnson said.

If a company doesn’t understand data science at all, it’s hard to hire for certain skills because hiring managers are unable to articulate what they need and why they need it.

“If they just want to do a stand-alone project, for example, they want a tool that optimizes scheduling for their workforce [which will take] a month or two of work to build the tool, then they won’t have much of a need for a full-time data scientist after that,” Johnson said.

A freelance data scientist can help decision-makers understand some of the basics, including what a data scientist does, what a data scientist needs to be successful and what data science can and cannot accomplish given the available data and other important factors that should be considered.

What can go wrong with contract help

If a company hires a full-time data scientist, most likely no one will expect that person to produce results on day one. Before a data scientist can share any valuable insights, that individual must first understand what the business hopes to achieve, what data is available, what data isn’t available, etc.

“The success of data science is completely predicated upon the data and if your data is insufficient, incomplete or inaccurate, you’re not going to get results — or good results — and the data scientist can’t fix that because the data you have is the data you have,” said Brandon Purcell, principal analyst at Forrester Research.

Nevertheless, unlike a new full-time data scientist, organizations often expect a freelance data scientist to be productive immediately just as with other types of contractors, and they struggle with getting results as quickly as desired.

“Even the most experienced data scientists face this problem as every company’s data can be extremely different,” said Robert O’Callaghan, director of data science at relationship commerce platform provider Ordergroove. O’Callaghan is also a former freelance data scientist.

“Unfortunately, that happens a lot of the time,” Purcell said. “A data scientist will come in and do their best, and they may be very talented but any model they create is only as good as a coin toss.”

Another misconception is that a freelance data scientist’s project is complete once the analysis is finished, when implementation and maintenance are also necessary for the company to extract business value from the data. For example, as new data comes in, a model must be tuned or it will drift, becoming less accurate.

“I have seen multiple brilliantly analyzed — and expensive — projects fail to deliver value due to businesses believing that a project was complete before the back-end work was in place,” O’Callaghan said. “[That’s] an issue that does not occur with full-time data scientists.”

It’s also important to understand what should happen after the contract concludes.

“In an ideal world you’d 100% plan ahead to say this data science freelancer will do this piece of work, and at the end of that I will have this insight and then I can do X, Y or Z,” O’Callaghan said. “You can never really 100% anticipate your results, so you will need to be more flexible in understanding what the next step is once the work is complete.”

Fundamentally, companies are not scoping freelance data science projects appropriately. And, they may underestimate the impact these insights will have on business operations.

“You’re going to be using that analysis to change the way you interact with customers, perform your operations or the way your human resources behave,” Purcell said. “That’s going to take longer than building a model. [If the analysis doesn’t result in] process changes [or] operational changes, there’s a good chance the model is going to end up being this shiny science project that never gets adopted.”

Bottom line

If you don’t already have a data science function, a freelance data scientist may help you better understand the opportunities and pitfalls. Freelancers are also a good choice for project work whether a data science function exists or not.

Be wary of making assumptions about what data science and data scientists can and can’t do if you don’t have the benefit of expert insight, however. Otherwise, your data science efforts and their results may fall short of expectations or fail entirely.

The post Should your company hire a freelance data scientist? appeared first on Artificial Intelligence.

Google AI has open-sourced A Lite Bert (ALBERT), a deep-learning natural language processing (NLP) model, which uses 89% fewer parameters than the state-of-the-art BERT model, with little loss of accuracy. The model can also be scaled-up to achieve new state-of-the-art performance on NLP benchmarks.

The research team described the model in a paper to be presented at the International Conference on Learning Representations. ALBERT uses two optimizations to reduce model size: a factorization of the embedding layer and parameter-sharing across the hidden layers of the network. Combining these two approaches results in a baseline model with only 12M parameters, compared to BERT’s 108M, while achieving an average of 80.1% accuracy on several NLP benchmarks compared with BERT’s 82.3% average. The team also trained a “double-extra-large” ALBERT model with 235M parameters which performed better on benchmarks than the “large” BERT model with 334M parameters.

Recent advances in state-of-the-art NLP models have come from pre-training large models on large bodies of unlabeled text data using “self-supervision” techniques. However, the large size of these models, with hundreds of millions of parameters, present an obstacle to experimentation. Not only does training time and cost go up with model size, at some point the models are simply too large to train; they cannot fit in the memory of the training computers. While there are techniques to address this, the Google AI team has identified ways to reduce model size without sacrificing accuracy. With smaller models, the researchers can better explore the hyperparameter space of the models:

in order to improve upon this new approach to NLP, one must develop an understanding of what, exactly, is contributing to language-understanding performance — the network’s height (i.e., number of layers), its width (size of the hidden layer representations), the learning criteria for self-supervision, or something else entirely?

The first of ALBERT’s optimizations is a factorization of the word embeddings. ALBERT, like BERT and many other deep-learning NLP models, is based on the Transformer architecture. The first step in this model is to convert words to numeric “one-hot” vector representations. The one-hot vectors are then projected into an embedding space. A restriction of the Transformer is that the embedding space must have the same dimension as the size of the hidden layers. Projecting a vocabulary of size V into an embedding of dimension E requires VxE parameters. With the large vocabularies and model dimensions needed to achieve state-of-the-art results, this could require close to a billion parameters. By factorizing the embedding, the ALBERT team first projects the word vectors into a smaller-dimensional space: 128 vs BERT’s 768. Then this smaller embedding is projected into a higher-dimensional space that has the same dimension as the hidden layers. The team posits that the first projection is a context-independent representation of the word, while the second is context-dependent.

The second optimization is to share parameters across the network’s layers. Transformer network layers contain both a feed-forward component and an attention component; ALBERT’s strategy is to share each component across all layers. This does result in a loss of accuracy of about 1.5 percentage points, but it does reduce the number of parameters needed from 89M to 12M.

Google has released a TensorFlow-based implementation of ALBERT as well as models trained on an English-language corpus and a Chinese-language corpus; users on Twitter are now asking if Google has plans to release a model trained on a Spanish-language corpus. The ALBERT code and models are available on GitHub.

The post Google Open-Sources ALBERT Natural Language Model appeared first on Artificial Intelligence.

We produce 2.5 exabytes of data every day. That’s equivalent to 250,000 Libraries of Congress or the content of 5 million laptops. Every minute of every day 3.2 billion global internet users continue to feed the data banks with 9,722 pins on Pinterest, 347,222 tweets, 4.2 million Facebook likes plus ALL the other data we create by taking pictures and videos, saving documents, opening accounts and more.

The post How Quantum Computers Will Revolutionize Artificial Intelligence, Machine Learning And Big Data appeared first on Artificial Intelligence.

Google Analytics now has the same natural language processing technology available in other Google apps such as Photos and Search, the Internet giant announced on Tuesday. That means you can now ask questions in plain English, which gets you answers quicker than before.

Depending on the question you ask, you will be presented with a number, rows, or a chart. As an example, the Analytics team suggests a question such as “How many new users did we have from organic search on mobile last week?” which will give you a number. If you ask the trend of session duration, expect the answer in a chart form as the image above.

The feature becomes a part of Analytics Intelligence, which relies on machine learning to make sense of your analytics data. Analytics Intelligence will also help provide automated insights – now available on both Web and the app – alongside smart lists, smart goals, and session quality. Insights will also present specific recommendations to improve metrics, such as reducing load time to decrease bounce rate, and adding a new AdWords keyword to boost conversion rate.

To access the new questions feature and get automated insights, click the Intelligence button to open a side panel on the website, and tap the Intelligence icon in the top-right corner on the Google Analytics app for Android and iOS.

Google says the new features are now rolling out, and will be available in English to all Google Analytics users over the next few weeks.

The post Google Analytics Gets Natural Language Processing Support With Machine Learning appeared first on Artificial Intelligence.