The Big Problems of Big Data
June 30, 2017
Companies are producing volumes of data. However, no fully functional system is able to provide actionable insights to decision makers in real time. Bayesian methods might pave the way to the solution seekers.
In an article published by PHYS and titled Advances in Bayesian Methods for Big Data, the author says:
Bayesian methods provide a principled theory for combining prior knowledge and uncertain evidence to make sophisticated inference of hidden factors and predictions.
Though the methods of data collection have improved, analyzing and presenting actionable insights in real time is still a big problem for Big Data adopters. Human intervention is required at almost every step which defies the entire purpose of an intelligent system. Hopefully, Bayesian methods can resolve these issues. Experts have been reluctant to adopt Bayesian methods owing to the fact that they are slow and are not scalable. However, with recent advancements in machine learning, the method might work.
Vishal Ingole, June 30, 2017
Bing Introduces an Image Feed
June 30, 2017
Here’s a short write-up about a notable addition to Bing —On MSFT reports, “Bing Image Search Updated with Image Feed, Taking on Pinterest.” After noting that the Tools menu has been renamed “Filter” and moved to the right of the screen, writer Jack Wilkinson explains:
A new feature has also appeared, known as Image Feed, which replaces where Tools originally used to be placed. Image Feed allows you to choose a feed of images…. When selecting an image feed to look at, it allows you to follow it as an ‘interest’, so that you can see new images in a feed. Your personalised image feed can be accessed here. By the looks of it, it appears as though Bing’s new image feed is taking a hit at Pinterest – bringing all the images you could want into one place via a feed, in similar fashion to Pinterest.
Yes, this could certainly replace Pinterest for many users, especially ones who already frequent Bing. I had noticed the refine-by-keyword list at the top of Google’s image results page is formatted much like the one on my Pinterest account. Will that online search platform, still number one by far, also implement a Pinterest-like image feed? Stay tuned.
Cynthia Murrell, June 30, 2017
Qwant Makes a Bold Prediction
June 29, 2017
Is there any platform that can rival Google at Internet search? Qwant believes it can, we learn from the article, “Qwant, a French Search Engine, Thinks it Can Take on Google—Here’s Why” at Search Engine Journal. Writer Matt Southern points to a Motherboard article to support his assertion, and relates:
Like search engine DuckDuckGo, Qwant’s competitive advantage is privacy. It protects users’ privacy by not tracking what they’re doing or searching for online. Qwant doesn’t use cookies, collect browsing data, or do any kind of data profiling.
So, other than privacy, what does Qwant do that sets itself apart from Google? Or even DuckDuckGo for that matter? For one, it currently has over 31 different search categories. In addition to the standard news, images, and video categories, Qwant offers categories such as: social media, music, jobs, cars, health, and more.
The company also has a unique philosophy that artificial intelligence and digital assistants can be educated without having to collect data on users. That’s a completely different philosophy than what is shared by Google, which collects every bit of information it can about users to fuel things like Google Home and Google Allo.
Naturally, Qwant needs to earn money, and it currently does so through click-throughs; the company also has plans to work with TripAdvisor and eBay down the line. Currently, users can make Qwant their default search engine within Firefox, and they hope to expand that to other browsers. Qwant backs up its privacy commitment by providing its source code to third-party data protection agencies. Launched in 2013, the company is based in Paris.
Cynthia Murrell, June 29, 2017
Alexa Is Deaf to Vocal Skill Search
June 29, 2017
Here is a surprising fact: Amazon does not have a vocal search for Alexa skills. Amazon prides itself on being on being a top technology developer and retailer, but it fails to allow Alexa users to search for a specific skill. Sure, it will list the top skills or the newest ones, but it does not allow you to ask for any specifics. Tech Crunch has the news story: “Amazon Rejects AI2’s Alexa Skill Voice-Search Engine. Will It Build One?”
The Allen Institute For Artificial Intelligence decided to take the task on themselves and built “Skill Search.” Skill Search works very simply: users state what they want and then Skill Search will list other skills that can fulfill the request. When AI2 submitted the Skill Search to Amazon it was rejected on the grounds that Amazon does not want “skills to recommend other skills.” This is a pretty common business practice for companies and Amazon did state on its policy page that skills of this nature were barred. Still, Amazon is missing an opportunity:
It would seem that having this kind of skill search engine would be advantageous to Amazon. It provides a discovery opportunity for skill developers looking to get more users, and highlighting the breadth of skills could make Alexa look more attractive compared to alternatives like Google Home that don’t have as well established of an ecosystem.
Amazon probably has a vocal search skill on their projects list and does not have enough information about it to share yet. Opening vocal search gives Amazon another revenue stream for Alexa. They are probably working on perfecting the skill’s language comprehension skills. Hey Amazon, maybe you should consider Bitext’s language offerings for an Alexa skills search?
Whitney Grace, June 29, 2017
Google, Microsoft Trail Amazon AWS Success
June 28, 2017
Is it now impossible for any fourth company to claw their way to the top of the cloud? Amazon’s AWS is still way ahead in the hosted-services game, but Google’s CloudPlatform and Microsoft’s Azure are also flourishing, we learn from “AWS, Google, and Microsoft Cement Their Cloud Dominance” at InfoWorld. Writer Matt Asay observes:
[AWS is] redefining enterprise IT forever, with everyone else having to sprint to catch up to its torrid pace of innovation and price cuts. Interestingly, Microsoft Azure and Google Cloud appear to be doing exactly that. … The market is consolidating around the big three cloud vendors, and it may be too late to add a fourth.
Though both Microsoft and Google have yet to break out their respective cloud revenue numbers (they’re not required to until it represents 5 percent of their overall company revenue), both reported big increases for 2016. Microsoft’s Azure revenue grew 95 percent in constant currency, and while Google didn’t report revenue or growth numbers, it suggested it had ‘significant momentum’ in the year, a claim made more believable by deals like Snap’s ($2 billion over five years for Google Cloud services). The problem for Microsoft and Google is that AWS isn’t slowing down. If anything, it’s accelerating.
Ah, lively competition. Asay emphasizes that much of Amazon’s success has occurred because they were able to reduce prices while increasing functionality. That is indeed a difficult combination to beat. Google or Microsoft may be in the position to pull off an upset, perhaps with some remarkable feat of innovation. Fourth place and below may be wise to pursue revenue streams outside the hosting field.
Cynthia Murrell, June 28, 2017
Google Translate Is Constantly Working
June 28, 2017
It seems we are always just around the corner from creating the universal translation device, but with online translation services the statement is quite accurate. Google Translate is one of the most powerful and accurate free translation services on the Internet. Lackuna shares some “Facts About Google Translate You May Not Know” to help you understand the power behind Google Translate.
Google Translation is built on statistical machine translation (SMT), basically it means that computers are analyzing translated documents from the Web to learn languages and find the patterns within them. From there, the service picks the best probable translation results for each query. Google Translate used to work differently:
However, Google Translate didn’t always work this way. Initially, it used a rule-based system, where rules of grammar and syntax, along with vocabulary for each language, were manually coded into a computer. Google switched to SMT because it enables the service to improve itself as it combs the web adding to its database of text — as opposed to linguists having to identify and code new rules as a language evolves. This provides a much more accurate translation, as well as saving thousands of programming/linguist man-hours.
While Google might be saving time relying fully on SMT, the linguist human touch is necessary to gain the sentimental and full comprehension of a language. Companies like Bitext that built analytics engines on linguistics’ knowledge combined with machine learning have a distinct advantage over others.
Meanwhile, Google Translate still remains a valuable service. It currently translates sixty-four languages, a chatbot translates in real-time and allows people to communicate in their native tongue, it has a speech-to-speech translation in conversation mode node for Android, and also uses photos to translate written language in real time.
Whitney Grace, June 28, 2017
Stephen E Arnold: Dark Web Use Expected to Increase
June 27, 2017
Author predicts filtering and other restrictions on the open Internet will push more users toward secret encrypted platforms
Despite stepped-up efforts by federal and local law enforcement agencies, the Dark Web and the contraband markets that thrive there will continue to grow in the coming years. That’s the conclusion shared by author and consultant Stephen E Arnold in his new book Dark Web Notebook, a practical guide for law enforcement, intelligence, and corporate security personnel.
Arnold was a featured speaker this month at the at the Techno Security & Digital Forensics Conference in Myrtle Beach, SC, where he explained how the Dark Web operates and how investigators can assume a secret identity and acquire the necessary tools to explore hidden online communities.
In his lectures, Arnold noted that governments in the UK, China, Egypt, and other countries are implementing tighter restrictions on electronic communications while private companies such as Facebook and Google are implementing more content filtering. The result, he said, is that more activities that are normally conducted on the “surface web” are migrating to the encrypted platforms offered by the Dark Web. At the same time, private citizens are adopting apps and other encryption tools to protect their personal privacy.
“It’s like squeezing a sponge or a tube of toothpaste,” Arnold said. “The substance has to go somewhere. The Dark Web will become more attractive to people who want to conduct their affairs in secret and that will make the Dark Web an increasing area of interest to law enforcement, security and intelligence professionals.”
The Dark Web is also flashing on the radar of state and local law enforcement agencies. A June 10 article in the New York Times (https://www.nytimes.com/2017/06/10/business/dealbook/opioid-dark-web-drug-overdose.html) described how police in the resort town of Park City, UT, had to learn about the Dark Web after two 13-year-old boys died after taking a synthetic opioid called “Pinky” that was purchased on the Dark Web.
“This terrible tragedy is an example of how the influence of the Dark web is spreading into small town in the US and around the world. Law enforcement personnel at all levels are going to need to understand how this new environment functions if they hope to protect the citizens in their communities,” Arnold said.
Although much has been written about the Dark Web, Arnold’s book is unique in that it was created specifically for investigators. It presents step-by-step instructions that help investigators safely explore the digital underworld. The book is available as a 250-page digital download for $49 at https://gum.co/darkweb.
In addition to his lectures, Arnold also offers a hands-on tutorial for small groups of qualified participants. In these sessions, participants use a secure computer to create a “legend” identity and explore the Dark We. They also learn how to acquire other digital tools and services including Bitcoin, the digital currency favored by Dark Web vendors.
Arnold will present his lectures and tutorial again at the 2017 Techno Security & Digital Forensics Conference September 18-20 in San Antonio. Information about the conference is available at www.technosecurity.us.
About the Book
The 250-page book includes a tour of the Dark Web with profiles of vendors and markets, explanations of access tools such as the Tor browser, reviews of services for searching Dark Web content, research resources available on the public Internet, and details about commercial solutions.
Detailed descriptions of each chapter are posted at http://xenky.com/darkwebnotebook.
The book is available as a digital download for $49 at https://gum.co/darkweb.
About the Author
Stephen E Arnold’s career includes work at Halliburton Nuclear Utility Services and Booz, Allen & Hamilton. He built and sold several stary-up ventures including The Point Internet Service. He and his team built the online system for the Threat Open Intelligence Gateway (TOSIG) for the US government. He is the author of The Google Legacy (2005), Google Version 2 (2007), and Google: The Digital Gutenberg (2009). In 2015, he published CyberOSINT: Next Generation Information Access. He is a Summa cum laude graduate of Bradley University and he has completed work on his PhD at the University of Illinois.
About ArnoldIT
Arnold Information Technology conducts research into online information and services. Founded by Stephen E Arnold in 1991, the company has provided technology services to organizations worldwide.
Ric Manning, June 27, 2017
HonkinNews for 27 June Now Available
June 27, 2017
Excitement abounds in search and content processing. In this, the 27 June 2017 edition of HonkinNews, we report on Booz Allen’s exciting public relations challenge. First there was Snowden, then there was a pack rat stuffing top secret documents in his Maryland home. Now, the US government is investigating the once prestigious firm for tripping over its green eye shades and shirt gaiters. Attivio continues to lead the pack of search and retrieval companies in marketing innovations. We run down the jargon the 10 year old company uses to sell its “kitchen sink” solution. In addition to the old chestnut “searh,” Attivio offers natural language processing and the IBM favorite “cognitive.” IBM has found a new source of management ideas. Yahoot’s Marissa Mayer decided telecommuting was not a good thing. She implemented changes in the work from home policy and sold the company to Verizon. Now IBM is following in Yahoot’s footsteps. We ask, “What other management ideas can IBM tap for business inspiration?” If you are a Google watcher, you may find out mini-feature on Google’s reinvention of Monster.com. The ad search giant may be less interested in helping an unemployed cigarette conveyer belt repairman find work. The goal of the new Jobs service may be Microsoft LinkedIn. Enjoy the show at this link.
Kenny Toth, June 27, 2017
Scadarlia Refines Internet Search Results
June 27, 2017
You can add a touch of arts-and-crafts to your online searches with a third-party preview-and-notation app—“Scadarlia: New Approach to Search Engines Using.” The promo page includes a video and is full of illustrative screenshots. What interests us is the way Scadarlia evaluates the relevance of each result. The Softpedia download page goes into the tech behind the folksy-looking add-on:
The program prompts you with a main window that is split into two sections, which should reinterpret your approach to search engines. While the left section is dedicated to keywords as well as the list of results the search engine considers suitable for your inquiry, the right panel shows the URL you want to analyze in detail. While this may look like a program packing ordinary browser-like capability, it is not. In fact, the application differentiates itself through its ability to follow a series of rules when displaying the results of a Google or Bing search. It can analyze the position of your keywords within your page, making sure that they are as close to one another as possible, since this is what makes them more representative for what you have in mind.
Other features include the color-coding sites by usefulness and the abilities to blacklist sites and to create stop words. The full version can be downloaded for $9.95 from its Softpedia page.
Cynthia Murrell, June 27, 2017
Apple Lovers Demand Their Own Talking Speaker
June 27, 2017
Google and Amazon dominate the intelligent speaker market and it is about to get more crowded. Marketing Land reports on a recent Morning Consult survey that showed how Apple lovers would like their own talking speaker: “Survey: Amazon Echo, Google Home Onwers ‘Very Interested’ In Apple HomePod.” Morning Consult surveyed 2,000 US consumers and discovered that one third of them are interested the Apple HomePod and 45 percent are Apple users.
Even more surprising among the results is that the consumers who are the most interested to use an Apple HomePod already own the competing devices. There are more interesting numbers:
According to the survey, the following were the rankings of variables, “among those who said [the] feature was ‘very important’ when considering a voice-controlled assistant:
57% Price
51% Speaker/audio quality
49% Accuracy of device’s voice recognition
44% Compatibility with devices you may already own, such as your smartphone
30% Access to a variety or music streaming services
29% Ability for device to integrate with other services or platforms, such as controlling smart light bulb
29% Brand that manufactures the device21% Aesthetics or look of the device
Is this an indicator that the Apple cult will win over the home digital assistant market? It might, but Amazon is still favored among consumers and might be the biggest contender because of the shopping connection and the price. The accuracy of the HomePod’s voice recognition is very important to consumers, especially when Siri fails to understand. Bitext could improve Apple, Google, and Amazons’ digital assistants when it comes to natural speech recognition.
Whitney Grace, June 27, 2017
-
- Subscribe to Beyond Search
Feature archive
News archive
-
