CBPO

Tag: Machine

Krisp reduces noise on calls using machine learning, and it’s coming to Windows soon

December 10, 2018 No Comments

If your luck is anything like mine, as soon as you jump on an important call, someone decides it’s a great time to blow some leaves off the sidewalk outside your window. 2Hz’s Krisp is a new desktop app that uses machine learning to subtract background noise like that, or crowds, or even crying kids — while keeping your voice intact. It’s already out for Macs and it’s coming to Windows soon.

I met the creators of Krisp, including 2Hz co-founder Davit Baghdasaryan, earlier this year at UC Berkeley’s Skydeck accelerator, where they demonstrated their then-prototype tech.

The tech involved is complex, but the idea is simple: If you create a machine learning system that understands what the human voice sounds like, on average, then it can listen to an audio signal and select only that part of it, cutting out a great deal of background noise.

Baghdasaryan, formerly of Twilio, originally wanted to create something that would run on mobile networks, so T-Mobile or whoever could tout built-in noise cancellation. This platform approach proved too slow, however, so they decided to go straight to consumers.

“Traction with customers was slow, and this was a problem for a young startup,” Baghdasaryan said in an email later. However, people were loving the idea of ‘muting noise,’ so we decided to switch all our focus and build a user-facing product.”

That was around the time I talked with them in person, incidentally, and just six months later they had released on Mac.

It’s simple: You run the app, and it modifies both the outgoing and incoming audio signals, with the normal noisy signal going in one end and a clean, voice-focused one coming out the other. Everything happens on-device and with very short latency (around 15 milliseconds), so there’s no cloud involved and nothing is ever sent to any server or even stored locally. The team is working on having the software adapt and learn on the fly, but it’s not implemented yet.

Another benefit of this approach is it doesn’t need any special tweaking to work with, say, Skype instead of Webex. Because it works at the level of the OS’s sound processing, whatever app you use just hears the Krisp-modified signal as if it were clean out of your mic.

They launched on Mac because they felt the early-adopter type was more likely to be on Apple’s platform, and the bet seems to have paid off. But a Windows version is coming soon — the exact date isn’t set, but expect it either late this month or early January. (We’ll let you know when it’s live.)

It should be more or less identical to the Mac version, but there will be a special gaming-focused one. Gamers, Baghdasaryan pointed out, are much more likely to have GPUs to run Krisp on, and also have a real need for clear communication (as a PUBG player I can speak to the annoyance of an open mic and clacky keys). So there will likely be a few power-user features specific to gamers, but it’s not set in stone yet.

You may wonder, as I did, why they weren’t going after chip manufacturers, perhaps to include Krisp as a tech built into a phone or computer’s audio processor.

In person, they suggested that this ultimately was also too slow and restrictive. Meanwhile, they saw that there was no real competition in the software space, which is massively easier to enter.

“All current noise cancellation solutions require multiple microphones and a special form factor where the mouth must be close to one of the mics. We have no such requirement,” Baghdasaryan explained. “We can do it with single-mic or operate on an audio stream coming from the network. This makes it possible to run the software in any environment you want (edge or network) and any direction (inbound or outbound).”

If you’re curious about the technical side of things — how it was done with one mic, or at low latency, and so on — there’s a nice explanation Baghdasaryan wrote for the Nvidia blog a little while back.

Furthermore, a proliferation of AI-focused chips that Krisp can run on easily means easy entry to the mobile and embedded space. “We have already successfully ported our DNN to NVIDIA GPUs, Intel CPU/GNA, and ARM. Qualcomm is in the pipeline,” noted Baghdasaryan.

To pursue this work the company has raised a total of $ 2 million so far: $ 500K from Skydeck as well as friends and family for a pre-seed round, then a $ 1.5 M round led by Sierra Ventures and Shanda Group.

Expect the Windows release later this winter, and if you’re already a user, expect a few new features to come your way in the same time scale. You can download Krisp for free here.

Gadgets – TechCrunch


TechSEO Boost: Machine Learning for SEOs

December 4, 2018 No Comments

This year’s TechSEO Boost, an event dedicated to technical SEO and hosted by Catalyst, took place on November 29 in Boston.

Billed as the conference “for developers and advanced SEO specialists,” TechSEO Boost built on the success of the inaugural event in 2017 with a day of enlightening, challenging talks from the sharpest minds in the industry.

Some topics permeated the discourse throughout the day and in particular, machine learning was a recurring theme.

As is the nature of the TechSEO Boost conference, the sessions aimed to go beyond the hype to define what precisely machine learning means for SEO, both today and in future.

The below is a recap of the excellent talk from Britney Muller, Senior SEO Scientist at Moz, entitled (fittingly enough) “Machine Learning for SEOs.”

What is machine learning? A quick recap.

The session opened with a brief primer on the key terms and concepts that fit under the umbrella of “machine learning.”

Muller used the definition in the image below to capture the sense of machine learning as “a subset of AI (Artificial Intelligence) that combines statistics and programming to give computers the ability to “learn” without explicitly being programmed.”

definition what is machine learning

That core idea of “learning” from new stimuli is an important one to grasp as we consider how machine learning can be applied to daily SEO tasks.

Machine learning excels at identifying patterns in huge quantities of data. As such, some of the common examples of machine learning applications today include:

  • Recommender systems (Netflix, Spotify)
  • Ridesharing apps (Uber, Lyft)
  • Digital Assistants (Amazon Alexa, Apple Siri, Google Assistant)

This very ubiquity can make it a challenging concept to grasp, however. In fact, Eric Schmidt at Google has gone so far as to say, “The core thing Google is working on is basically machine learning.”

It is helpful to break this down into the steps that comprise a typical machine learning project, in order to see how we might apply this to everyday SEO tasks.

The machine learning process

The image below represents the machine learning process Muller shared at TechSEO Boost:

the machine learning process

It is important to bear in mind that some of the training data should be reserved for testing at a later point in the process.

Where possible, this data should also be labelled clearly to help the machine learning algorithm identify classifications and categories within a noisy data set.

It is for precisely this reason that Google asks us to label images to verify our identity:

street signs Google images

This demonstrates our human ability to pick out objects in cluttered contexts, but it has the added benefit of providing Google with higher quality image data.

The pitfalls of an unsupervised approach to machine learning, and a training data set that is open to interpretation, were laid bare just last week.

Google’s ‘Smart Compose’ feature within Gmail has demonstrated gender bias by preferring certain pronouns when predicting what a user might want to say.

As reported in Reuters, “Gmail product manager Paul Lambert said a company research scientist discovered the problem in January when he typed “I am meeting an investor next week,” and Smart Compose suggested a possible follow-up question: “Do you want to meet him?” instead of “her.”

gmail smart compose

The challenge here is not restricted to projects on such a scale. Marketers who want to get their hands dirty must be aware of the limitations of machine learning, as well as its exciting possibilities.

Muller added that people tend to overfit their data, which reduces the accuracy and flexibility of the model they are using. This (very common) phenomenon occurs when a model corresponds very closely with one specific data set, reducing its applicability to new scenarios.

The ability to scale effectively is what gives machine learning its appeal, so overfitting is something to be avoided with care. There is a good primer to this topic here and it is also explained very well through this image:

the best way to explain overfitting

So, how exactly can this subset of AI be used to improve SEO performance?

How you can use machine learning for SEO

As is the case with all hype-friendly technologies, businesses are keen to get involved with machine learning. However, the point is not to “use machine learning” through fear of being left behind, but rather to find the best uses of machine learning for each business.

Britney Muller shared some examples from her role at Moz during her session at TechSEO Boost.

The first was an approach to automated meta description generation using the Algorithmia Advanced Content Summarizer, which was then compared to Google’s approach to automated descriptions pulled directly from the landing page.

Meta descriptions remain an important asset when trying to encourage a positive click-through rate, but a lot of time is spent crafting these snippets. An automated alternative that can interpret the meaning of landing pages and create clickable summaries for display in the SERPs would be very useful.

serp q&a results comparison

Muller shared some examples, such as the image above, to demonstrate the comparison between the two approaches. The machine learning approach is not perfect and may require some tweaking, but it does an excellent job of conveying the page’s intent when compared to Google’s selection.

The team at Moz has since built this into Google Sheets:

google sheets meta descriptions

Although this is not a product other businesses can access right now, an alternative way of achieving automated meta descriptions has been shared by Paul Shapiro (the TechSEO Boost host) via Github here.

Automated image optimization

Another fascinating use of machine learning for SEO is the automation of image optimization. Britney Muller showed how, in under 20 minutes, it is possible to train an algorithm to distinguish between cats and ducks, then use this model on a new data set with a high level of accuracy.

recognize ducks vs snakes

For large retailers, the application of this method could be very beneficial. With so many new images added to the inventory every day, and with visual search on the rise, a scalable image labeling system would prove very profitable. As demonstrated at TechSEO Boost, this is now a very realistic possibility for businesses willing to build their own model.

A further use of machine learning described by Britney Muller was the transcription of podcasts. An automated approach to this task can turn audio files into something much more legible for a search engine, thereby helping with indexation and ranking for relevant topics.

Muller detailed an approach using the Amazon Transcribe product through Amazon Web Services to achieve this aim.

amazon transcribe tool

The audio is broken down and delivered in a J-SON file in a lot of detail, with the different speakers on the podcast labelled separately.

transcription json string

There was not enough time in the session to work through every potential use of machine learning for SEO, but Muller’s core message was that everyone in the industry should be working towards at least a working knowledge of these concepts.

Some further opportunities for experimentation were listed as follows:

opportunities with machine learning

As we can see, machine learning truly excels when working with large data sets to identify patterns.

Tools and resources

The best way to get engaged is to combine theory with practice. This is almost always the case, but it is a particularly valid piece of advice in relation to programming.

Muller’s was not the first or last talk to reference Google Codelabs throughout the day.

google codelabs

There are more resources out there than ever before and the likes of Amazon and Google want machine learning to be approachable. Amazon has launched a machine learning course and Google’s crash course is a fantastic way to learn the components of a successful project.

google codelabs machine learning crash course

The Google-owned Kaggle is always a great place to trial new data sets and review the innovative work performed by data scientists around the world, once a basic grasp has been attained.

Furthermore, Google’s Colaboratory makes it easy to get started on a project and work with a remote team.

Key takeaways: machine learning for SEOs

What became particularly clear through Muller’s talk is how approachable machine learning applications can be for SEOs. Moreover, the room for experimentation is unprecedented, for those willing to invest some time in the discipline.

key takeaways

The post TechSEO Boost: Machine Learning for SEOs appeared first on Search Engine Watch.

Search Engine Watch


Not Just Another AI and Machine Learning Discussion [Submit Your Questions to the Pros!]

November 30, 2018 No Comments

Time to talk through what you ACTUALLY want to know about AI and machine learning. Submit your questions!

Read more at PPCHero.com
PPC Hero


New Uber feature uses machine learning to sort business and personal rides

August 14, 2018 No Comments

Uber announced a new program today called Profile Recommendations that takes advantage of machine intelligence to reduce user error when switching between personal and business accounts.

It’s not unusual for a person to have both types of accounts. When you’re out and about, it’s easy to forget to switch between them when appropriate. Uber wants to help by recommending the correct one.

“Using machine learning, Uber can predict which profile and corresponding payment method an employee should be using, and make the appropriate recommendation,” Ronnie Gurion, GM and Global Head of Uber for Business wrote in a blog post announcing the new feature.

Uber has been analyzing a dizzying amount of trip data for so long, it can now (mostly) understand the purpose of a given trip based on the details of your request. While it’s certainly not perfect because it’s not always obvious what the purpose is, Uber believes it can determine the correct intention 80 percent of the time. For that remaining 20 percent, when it doesn’t get it right, Uber is hoping to simplify corrections too.

Photo: Uber

Business users can now also assign trip reviewers — managers or other employees who understand the employee’s usage patterns, and can flag questionable rides. Instead of starting an email thread or complicated bureaucratic process to resolve an issue, the employee can now see these flagged rides and resolve them right in the app. “This new feature not only saves the employee’s and administrator’s time, but it also cuts down on delays associated with closing out reports,” Gurion wrote in the blog post announcement.

Uber also announced that it’s supporting a slew of new expense reporting software to simplify integration with these systems. They currently have integrations with Certify, Chrome River, Concur and Expensify. They will be adding support for Expensya, Happay, Rydoo, Zeno by Serko and Zoho Expense starting in September.

All of this should help business account holders deal with Uber expenses more efficiently, while integrating with many of the leading expense programs to move data smoothly from Uber to a company’s regular record-keeping systems.


Enterprise – TechCrunch


Putting machine learning into the hands of every advertiser

July 31, 2018 No Comments


This post originally appeared on the Inside AdWords blog

The ways people get things done are constantly changing, from finding the closest coffee shop to organizing family photos. Earlier this year, we explored how machine learning is being used to improve our consumer products and help people get stuff done.

In just one hour, we’ll share how we’re helping marketers unlock more opportunities for their businesses with our largest deployment of machine learning in ads. We’ll explore how this technology works in our products and why it’s key to delivering the helpful and frictionless experiences consumers expect from brands.

Join us live today at 9am PT (12pm ET).

Deliver more relevance with responsive search ads

Consumers today are more curious, more demanding, and they expect to get things done faster because of mobile. As a result, they expect your ads to be helpful and personalized. Doing this isn’t easy, especially at scale. That’s why we’re introducing responsive search ads. Responsive search ads combine your creativity with the power of Google’s machine learning to help you deliver relevant, valuable ads.

Simply provide up to 15 headlines and 4 description lines, and Google will do the rest. By testing different combinations, Google learns which ad creative performs best for any search query. So people searching for the same thing might see different ads based on context.

We know this kind of optimization works: on average, advertisers who use Google’s machine learning to test multiple creative see up to 15 percent more clicks.1 Responsive search ads will start rolling out to advertisers over the next several months.

Maximize relevance and performance on YouTube

People watch over 1 billion hours of video on YouTube every day. And increasingly, they’re tuning in for inspiration and information on purchases large and small. For example, nearly 1 in 2 car buyers say they turn to YouTube for information before their purchase.2 And nearly 1 in 2 millennials go there for food preparation tips before deciding what ingredients to buy.3 That means it’s critical your video ads show at the right moment to the right audience.

Machine learning helps us turn that attention into results on YouTube. In the past, we’ve helped you optimize campaigns for views and impressions. Later this year, we’re rolling out Maximize lift to help you reach people who are most likely to consider your brand after seeing a video ad. This new Smart Bidding strategy is also powered by machine learning. It automatically adjusts bids at auction time to maximize the impact your video ads have on brand perception throughout the consumer journey.

Maximize lift is available now as a beta and will roll out to advertisers globally later this year.

Drive more foot traffic with Local campaigns

Whether they start their research on YouTube or Google, people still make the majority of their purchases in physical stores. In fact, mobile searches for “near me” have grown over 3X in the past two years4, and almost 80 percent of shoppers will go in store when there’s an item they want immediately.5 For many of you, that means driving foot traffic to your brick-and-mortar locations is critical—especially during key moments in the year, like in-store events or promotions.

Today we’re introducing Local campaigns: a new campaign type designed to drive store visits exclusively. Provide a few simple things—like your business locations and ad creative—and Google automatically optimizes your ads across properties to bring more customers into your store.

Show your business locations across Google properties and networks

Local campaigns will roll out to advertisers globally over the coming months.

Get the most from your Shopping campaigns

Earlier this year, we rolled out a new Shopping campaign type that optimizes performance based on your goals. These Smart Shopping campaign help you hit your revenue goals without the need to manually manage and bid to individual products. In the coming months, we’re improving them to optimize across multiple business goals.

Beyond maximize conversion value, you’ll also be able to select store visits or new customers as goals. Machine learning factors in the likelihood that a click will result in any of these outcomes and helps adjust bids accordingly.

Machine learning is also used to optimize where your Shopping ads show—on Google.com, Image Search, YouTube and millions of sites and apps across the web—and which products are featured. It takes into account a wide range of signals, like seasonal demand and pricing. Brands like GittiGidiyor, an eBay company, are using Smart Shopping campaigns to simplify how they manage their ads and deliver better results. GittiGidiyor was able to increase return on ad spend by 28 percent and drive 4 percent more sales, while saving time managing campaigns.

We’re also adding support for leading e-commerce platforms to help simplify campaign management. In the coming weeks, you’ll be able to set up and manage Smart Shopping campaigns right from Shopify, in addition to Google Ads.

Tune in to see more

This is an important moment for marketers and we’re excited to be on this journey with you. Tune in at 9am PT (12pm ET) today to see it all unfold at Google Marketing Live.

For the latest news, follow the new Google Ads blog. And check out g.co/adsannouncements for more information about product updates and announcements.

1 Internal Google data.
2 Google / Kantar TNS, Auto CB Gearshift Study, US, 2017. n=312 new car buyers who watched online video.
3 Google / Ipsos, US, November 2017.
4 Internal Google data, U.S., July–Dec. 2015 vs. July–Dec. 2017.
5 Google/Ipsos, U.S., “Shopping Tracker,” Online survey, n=3,613 online Americans 13+ who shopped in the past two days, Oct.–Dec. 2017.


Google Analytics Blog


Workday acquires Rallyteam to fuel machine learning efforts

June 11, 2018 No Comments

Sometimes you acquire a company for the assets and sometimes you do it for the talent. Today Workday announced it was buying Rallyteam, a San Francisco startup that helps companies keep talented employees by matching them with more challenging opportunities in-house.

The companies did not share the purchase price or the number of Rallyteam employees who would be joining Workday .

In this case, Workday appears to be acquiring the talent. It wants to take the Rallyteam team and incorporate it into the company’s engineering unit to beef up its machine learning efforts, while taking advantage of the expertise it has built up over the years connecting employees with interesting internal projects.

“With Rallyteam, we gain incredible team members who created a talent mobility platform that uses machine learning to help companies better understand and optimize their workforces by matching a worker’s interests, skills and connections with relevant jobs, projects, tasks and people,” Workday’s Cristina Goldt wrote in a blog post announcing the acquisition.

Rallyteam, which was founded in 2013, and launched at TechCrunch Disrupt San Francisco in September 2014, helps employees find interesting internal projects that might otherwise get outsourced. “I knew there were opportunities that existed [internally] because as a manager, I was constantly outsourcing projects even though I knew there had to be people in the company that could solve this problem,” Rallyteam’s Huan Ho told TechCrunch’s Frederic Lardinois at the launch. Rallyteam was a service designed to solve this issue.

Last fall the company raised $ 8.6 million led by Norwest Ventures with participation from Storm Ventures, Cornerstone OnDemand and Wilson Sonsini.

Workday provides a SaaS platform for human resources and finance, so the Rallyteam approach fits nicely within the scope of the Workday business. This is the 10th acquisition for Workday and the second this year.

Chart: Crunchbase

Workday raised over $ 230 million before going public in 2012.


Startups – TechCrunch


Workday acquires Rallyteam to fuel machine learning efforts

June 9, 2018 No Comments

Sometimes you acquire a company for the assets and sometimes you do it for the talent. Today Workday announced it was buying Rallyteam, a San Francisco startup that helps companies keep talented employees by matching them with more challenging opportunities in-house.

The companies did not share the purchase price or the number of Rallyteam employees who would be joining Workday .

In this case, Workday appears to be acquiring the talent. It wants to take the Rallyteam team and incorporate it into the company’s engineering unit to beef up its machine learning efforts, while taking advantage of the expertise it has built up over the years connecting employees with interesting internal projects.

“With Rallyteam, we gain incredible team members who created a talent mobility platform that uses machine learning to help companies better understand and optimize their workforces by matching a worker’s interests, skills and connections with relevant jobs, projects, tasks and people,” Workday’s Cristina Goldt wrote in a blog post announcing the acquisition.

Rallyteam, which was founded in 2013, and launched at TechCrunch Disrupt San Francisco in September 2014, helps employees find interesting internal projects that might otherwise get outsourced. “I knew there were opportunities that existed [internally] because as a manager, I was constantly outsourcing projects even though I knew there had to be people in the company that could solve this problem,” Rallyteam’s Huan Ho told TechCrunch’s Frederic Lardinois at the launch. Rallyteam was a service designed to solve this issue.

Last fall the company raised $ 8.6 million led by Norwest Ventures with participation from Storm Ventures, Cornerstone OnDemand and Wilson Sonsini.

Workday provides a SaaS platform for human resources and finance, so the Rallyteam approach fits nicely within the scope of the Workday business. This is the 10th acquisition for Workday and the second this year.

Chart: Crunchbase

Workday raised over $ 230 million before going public in 2012.


Enterprise – TechCrunch


Google Kubeflow, machine learning for Kubernetes, begins to take shape

May 5, 2018 No Comments

Ever since Google created Kubernetes as an open source container orchestration tool, it has seen it blossom in ways it might never have imagined. As the project gains in popularity, we are seeing many adjunct programs develop. Today, Google announced the release of version 0.1 of the Kubeflow open source tool, which is designed to bring machine learning to Kubernetes containers.

While Google has long since moved Kubernetes into the Cloud Native Computing Foundation, it continues to be actively involved, and Kubeflow is one manifestation of that. The project was only first announced at the end of last year at Kubecon in Austin, but it is beginning to gain some momentum.

David Aronchick, who runs Kubeflow for Google, led the Kubernetes team for 2.5 years before moving to Kubeflow. He says the idea behind the project is to enable data scientists to take advantage of running machine learning jobs on Kubernetes clusters. Kubeflow lets machine learning teams take existing jobs and simply attach them to a cluster without a lot of adapting.

With today’s announcement, the project begins to move ahead, and according to a blog post announcing the milestone, brings a new level of stability, while adding a slew of new features that the community has been requesting. These include Jupyter Hub for collaborative and interactive training on machine learning jobs and Tensorflow training and hosting support, among other elements.

Aronchick emphasizes that as an open source project you can bring whatever tools you like, and you are not limited to Tensorflow, despite the fact that this early version release does include support for Google’s machine learning tools. You can expect additional tool support as the project develops further.

In just over 4 months since the original announcement, the community has grown quickly with over 70 contributors, over 20 contributing organizations along with over 700 commits in 15 repositories. You can expect the next version, 0.2, sometime this summer.


Enterprise – TechCrunch


Artificial intelligence and machine learning: What are the opportunities for search marketers?

January 2, 2018 No Comments

Did you know that by 2020 the digital universe will consist of 44 zettabytes of data (source: IDC), but that the human brain can only process the equivalent of 1 million gigabytes of memory?

The explosion of big data has meant that humans simply have too much data to understand and handle daily.

For search, content and digital marketers to make the most out the valuable insights that data can provide, it is essential to utilize artificial intelligence (AI) applications, machine learning algorithms and deep learning to move the needle of marketing performance in 2018.

In this article, I will explain the advancements and differences between artificial intelligence (AI), machine learning and deep learning while sharing some tips on how SEO, content and digital marketers can make the most of the insights – especially from deep learning – that these technologies bring to the search marketing table.

I studied artificial intelligence in college and after graduating took a job in the field. It was an exciting time, but our programming capabilities, when looking back now, were rudimentary. More than intelligence, it was algorithms and rules that did their best to mimic how intelligence solves problems with best-guess recommendations.

Fast forward to today and things have evolved significantly.

The Big Bang: The big data explosion and the birth of AI

Since 1956, AI pioneers have been dreaming of a world where complex machines possess the same characteristics as human intelligence.

In 1996, the industry reached a major milestone when the IBM’s Deep Blue computer defeated a chess grandmaster by considering 200,000,000 chessboard patterns a second to make optimal moves.

Between 2000 and 2017, there were many developments that enabled great leaps forward. Most important were the geometric increases in the amount data collected, stored, and made retrievable. That mountain of data, which came to be known as big data, ushered in the advent of AI.

And it keeps growing exponentially: in 2016 IBM estimated that 90% of the world’s data had been generated over the last few years.

When thinking about AI, machine learning and deep learning, I find it helps to simplify and visualize how the 3 categories work and relate to each other –  this framework also works from a chronological, sub-set development and size perspective.

Artificial intelligence is the science of making machines do things requiring human intelligence. It is human intelligence in machine format where computer programs develop data-based decisions and perform tasks normally performed by  humans.

Machine learning takes artificial intelligence a step further in the sense that algorithms are programmed to learn and improve without the need for human data input and reprogramming.

Machine learning can be applied to many different problems and data sets. Google’s RankBrain algorithm is a great example of machine learning that evaluates the intent and context of each search query, rather than just delivering results based on programmed rules about keyword matching and other factors.

Deep learning is a more detailed algorithmic approach, taken from machine learning, that uses techniques based on logic and exposing data to neural networks (think human brain) so that the technology trains itself to perform tasks such as speech and image recognition.

Massive data sets are combined with pattern recognition capabilities to automatically make decisions, find patterns, emulate previous decisions, etc. Self-learning comes from here as the machine gets better from the more data that it is supplied.

Driverless cars, Netflix movie recommendations and IBMs Watson are all great examples of deep learning applications that break down tasks to make machine actions and assists possible.

Organic search, content and digital performance: Challenge and opportunity

Organic search (SEO) drives 51% of all website traffic and hence in this section it is only natural to explain the key benefits that deep-learning brings to SEO and digital marketers.

Organic search is a data-intensive business. Companies value and want their content to be visible on thousands or even millions of keywords in one to dozens of languages. Search best practices involve about 20 elements of on-page and off-page tactics. The SERPs themselves now come in more than 15 layout varieties.

Organic search is your market-wide voice of the customer, telling you what customers want at scale. However, marketers are faced with the challenge of making sense of so much data, having limited resources to mine insights and then actually act on the right and relevant insight for their business.

To succeed in highly demanding markets against your competitors’ many brands now requires the expertise of an experienced data analyst, and this is where machine learning and deep learning layers help recommend optimizations to content.

Connecting the dots with deep learning: Data and machine learning

The size of the organic data and the number of potential patterns that exist on that data make it a perfect candidate for deep learning applications. Unlike simple machine learning, deep-learning works better when it can analyse a massive amount of relevant data over long periods of time.

Deep learning and its ability to identify or prioritize material changes in interests and consumption behavior allows organic search marketers to gain a competitive advantage, be at the forefront of their industry, and produce the material that people need before their competitors, boosting their reputation.

In this way, marketers can begin to understand the strategies put forth by their competitors. They will see how well they perform compared to others in their industry and can then adjust their strategies to address the strengths or weaknesses that they find.

  • The insights derived from deep learning technologies blend the best of search marketing and content marketing practices to power the development, activation, and automated optimization of smart content, content that is self-aware and self-adjusting, improving content discovery and engagement across all digital marketing channels.
  • Intent data offers in-the-moment context on where customers want to go and what they want to know, do, or buy. Organic search data is the critical raw material that helps you discover consumer patterns, new market opportunities, and competitive threats.
  • Deep learning is particularly important in search, where data is copious and incredibly dynamic. Identifying patterns in data in real-time makes deep learning your best first defense in understanding customer, competitor, or market changes – so that you can immediately turn these insights into a plan to win.

To propel content and organic search success in 2018 marketers should let the machines does more of the leg work to provide the insights and recommendations that allow marketers to focus on the creation of smart content.

Below are a just a few examples of the benefits for the organic search marketer:

Site analysis

Pinpoint and fix critical site errors that drive the greatest benefits to a brand’s bottom line. Deep learning technology can be used to incorporate website data, detect anomalies tying site errors to estimated marketing impact so that marketers can prioritize fixes for maximum results.

Without a deep learning application to help you, you might be staring at a long list of potential fixes which typically get postponed to later.

Competitive strategy

Identifying patterns in real-time makes deep learning a brands’ best first defense in understanding customer, competitor, or market changes– so that marketers can immediately turn these insights into a plan to win.

Content discovery

Surface high-value topics that target different content strategies, such as stopping competitive threats or capitalizing on local demand.

Deep learning technology can be used to assess the ROI of new content items and prioritize their development by unveiling insights such as topic opportunity, consumer intent, characteristics of top competing content, and recommendations for improving content performance.

Content development

Score the quality and relevance of each piece of content produced. Deep learning technology can help save time with automated tasks of content production, such as header tags, cross-linking, copy optimization, image editing, highly optimized CTAs that drive performance, and embedded performance tracking of website traffic and conversion.

Content activation

Deep learning technology can help ensure that each piece of content is optimized for organic performance and customer experience—such as schema for structure, AMP for better mobile experiences, and Open Graph for Facebook. Technology can help marketers can amplify their content in social networks for greater visibility.

Automation

Automation helps marketers do more with less and execute more quickly. It allows marketers to manage routine tasks with little effort, so that they can focus on high-impact activities and accomplish organic business goals at scale.

Note: To make the most of the insights and recommendations from deep learning marketers need to take action and make the relevant changes to web page content to keep website visitors engaged and ultimately converting.

Additionally, because the search landscape changes so frequently, deep learning fuels the development of smart content and can be used to automatically adjust to changes in content formats and standards.

Deep learning in action

An example of deep learning in organic search is DataMind. BrightEdge (disclosure, my employer) Data Mind is like a virtual team of data scientists built into the platform, that combines massive volumes of data with immediate, actionable insights to inform marketing decisions.

In this case the deep learning engine analyzes huge, complex, and dynamic data sets (from multiple sources that include 1st and 3rd party data) to determine patterns and derive the insights marketers need. Deep learning is used to detect anomalies in a site’s performance and interpret the reasons, such as industry trends, while making recommendations about how to proceed.

Conclusion

Think of deep learning applications as your own personal data scientist – here to help and assist and not to replace. The adoption of AI, machine learning and now deep learning technologies allows faster decisions, more accurate and smarter insights.

Brands compete in the content battleground to ensure their content is optimized and found, engages audiences and ultimately drives conversions and digital revenue. When armed with these insights from deep learning, marketers get a new competitive weapon and a massive competitive edge.

Search Engine Watch


Facebook finishes its move to neural machine translation

August 5, 2017 No Comments

 Facebook announced this morning that it had completed its move to neural machine translation — a complicated way of saying that Facebook is now using convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to automatically translate content across Facebook. Read More
Enterprise – TechCrunch