Google Antitrust Round Up
December 23, 2015
I read “Google Alphabet Can’t Just Sing A-B-C As EU Antitrust Chief Hints Of Fresh Competition Charges.” The write provides a summary of the antitrust issues swirling around Google. One investigation caught my attention. The write up states:
Last month, it was brought to light that India’s competition regulator, the Competition Commission of India, accused Google Search of generating biased search results which favored not only its own offerings but also those of companies that pay to advertise on its search engine service.
The Alphabet Google construct has lofty plans for India. The Loon balloon project is intended to lift revenues in that densely populated nation state.
Worth watching the Alphabet Google take off in India.
Stephen E Arnold, December 23, 2015
GoDaddy and Search
December 23, 2015
Years ago, I understood that GoDaddy, the domain name outfit, purchased a search company doing business as Innerprise. The press release issued in September 2004 said:
[GoDaddy] will incorporate Innerprise’s search products, including Enterprise Search 2004 and Innerprise Hosted Search, into the GoDaddy product catalog, augmenting the Company’s complete line of Web development tools including domain name registration, hosting, email systems, SSL certificates, and other complementary products and services that assist customers in building and maintaining a presence on the Internet.
I learned in “Why GoDaddy Built Its Search Engine from Scratch”:
GoDaddy, seeking to improve customer service, built a custom search engine that generates domain names on the fly for its small business customers. Building it wasn’t the best option, the company’s executives say. It was the only option.
The write up points out:
Custom software development is the preferred approach among online businesses.
How did GoDaddy meet its need for a search system:
… The engineering feat required GoDaddy to create search crawlers that can traverse hundreds of international registries, including in South Africa and Indonesia, generating tens of thousands of potential domain names in near real-time. The company also built machine-learning algorithms, in conjunction with open source Hadoop data processing software, to help surface the best domain names it can.
The write up does not reference the Innerprise solution. There is no hint of the cost of the system. The message is that an enterprising Yahoo alum can build a search engine from scratch, and you should too. There’s a new project for your New Year’s resolutions list.
Stephen E Arnold, December 23, 2015
RankBrain, the Latest AI from Google, Improves Search Through Understanding and Learning
December 23, 2015
The article on Entrepreneur titled Meet RankBrain, the New AI Behind Google’s Search Results introduces the AI that Google believes will aid the search engine in better understanding the queries it receives. RankBrain is capable of connecting related words to the search terms based on context and relevance. The article explains,
“The real intention of this AI wasn’t to change visitors’ search engine results pages (SERPs) — rather, it was to predict them. As a machine-learning system, RankBrain actually teaches itself how to do something instead of needing a human to program it…According to Jack Clark, writing for Bloomberg on the topic: “[Rankbrain] uses artificial intelligence to embed vast amounts of written language into mathematical entities — called vectors — that the computer can understand.”
Google scientist Greg Corrado spoke of RankBrain actually exceeding his expectations. In one experiment, RankBrain beat a team of search engineers in predicting which pages would rank highest. (The engineers were right 70% of the time, RankBrain 80%.) The article also addresses concerns that many vulnerable brands relying on SEOs may have. The article ventures to guess that it will be mainly newer brands and services that will see a ranking shift. But of course, with impending updates, that may change.
Chelsea Kerwin, December 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Google Clamps down on Surprise Costs in BigQuery
December 23, 2015
The article titled Google Promises to Rein in Runaway Query Costs on Fortune discusses the obstacles facing Google’s BigQuery data tool. Google hopes to make BigQuery a major resource for big companies considering cloud technology, but unpredictable costs are getting in the way of the “low-cost big data analytics option” marketing that Google has deployed. Hence, the introduction of “custom quota” and Query Explain,
“Google is now offering potential inquisitors a way to set a “custom quota” to ensure that the number crunching on a specified project does not exceed a pre-set daily limit. In addition, a Query Explain feature promises to lay out, how BigQuery will go about processing the question on the table in advance. That way, in theory, you can see if your questions will be “write, read, or compute heavy” and better anticipate where performance bottlenecks could lurk…”
One might fairly ask why there was any delay in these services, since customers are not known for their fondness of mobile phone type billing surprises. Amazon is also standing next to Google waving at RedShift, a BigQuery competitor in the air. But the simpler pricing and efficiency of BigQuery might be more appealing to many companies, especially with the more controlled processes now available.
Chelsea Kerwin, December 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Top Trends for Cyber Security and Analytics in 2016
December 23, 2015
With the end of the year approaching, people try to predict what will happen in the New Year. The New Year brings on a sort of fortunetelling, because if companies are able to correctly predict what will happen in 2016 then it serves for positive profit margins and a healthier customer base. The IT industry has its own share of New Year soothsayers and the Executive Biz blog shares that “Booz Allen Cites Top Cyber, Analytics Trends In 2016; Bill Stewart Comments” with possible trends in cyber security and data analytics for the coming year.
Booz Allen Hamilton says that companies will want to merge analytical programs with security programs to receive data sets that show network vulnerabilities; they have been dubbed “fusion centers.”
“ ‘As cyber risk and advanced analytics demand increasing attention from the C-suite, we are about to enter a fundamentally different period,’ said Bill Stewart, executive vice president and leader of commercial cyber business at Booz Allen. ‘The dynamics will change… Skilled leaders will factor these changing dynamics into their planning, investments and operations.’”
The will also be increased risks coming from the Dark Web and risks that are associated with connected systems, such as cloud storage. Booz Allen also hints that companies will need skilled professionals who know how to harness cyber security risks and analytics. That suggestion is not new, as it has been discussed since 2014. While the threat from the Internet and vulnerabilities within systems has increased, the need for experts in these areas as well as better programs to handle them has always been needed. Booz Allen is restating the obvious, the biggest problem is that companies are not aware of these risks and they usually lack the budget to implement preemptive measures.
Whitney Grace, December 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
The Ins and Outs of Hacking Software
December 23, 2015
Hacking software is and could be a potential problem. While some government agencies, hacktivist organizations, and software companies are trying to use it for good, terrorist groups, digital thieves, and even law enforcement agencies can use it to spy and steal data from individuals. The Technology Review shares some interesting stories about how software is being used for benign and harmful purposes in “The Growth Industry Helping Governments Hack Terrorists, Criminals, And Political Opponents.”
The company Hacking Team is discussed at length and its Remote Control System software, which can worm its way through security holes in a device and steal valuable information. Governments from around the globe have used the software for crime deterrence and to keep tabs on enemies, but other entities used the software for harmful acts including spying and hacking into political opponents computers.
Within the United States, it is illegal to use a Remote Control System without proper authority, but often this happens:
“When police get access to new surveillance technologies, they are often quickly deployed before any sort of oversight is in place to regulate their use. In the United States, the abuse of Stingrays—devices that sweep up information from cell phones in given area—has become common. For example, the sheriff of San Bernardino County, near Los Angeles, deployed them over 300 times without a warrant in the space of less than two years. That problem is only being addressed now, years after it emerged, with the FBI now requiring a warrant to use Stingrays, and efforts underway to force local law enforcement to do the same. It’s easy to imagine a similar pattern of abuse with hacking tools, which are far more powerful and invasive than other surveillance technologies that police currently use.”
It is scary how the software is being used and how governments are skirting around its own laws to use it. It reminds me of how gun control is always controversial topic. Whenever there is a mass shooting, debates rage about how the shooting would never had happened if there was stricter gun control to keep weapons out of the hands of psychopaths. While the shooter was blamed for the incident, people also place a lot of blame on the gun, as if it was more responsible. As spying, control, and other software becomes more powerful and ingrained in our lives, I imagine there will be debates about “software control” and determining who has the right to use certain programs.
Whitney Grace, December 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Machine Learning Used to Decipher Lute Tablature
December 23, 2015
The Oxford Journal’s Early Music publication reveals a very specialized use of machine learning in, “Bring ‘Musicque into the Tableture’: Machine-Learning Models for Polyphonic Transcription of 16th-Century Lute Tablature” by musical researchers Reinier de Valk and Tillman Weyde. Note that this link will take you to the article’s abstract; to see the full piece, you’ll have to subscribe to the site. The abstract summarizes:
“A large corpus of music written in lute tablature, spanning some three-and-a-half centuries, has survived. This music has so far escaped systematic musicological research because of its notational format. Being a practical instruction for the player, tablature reveals very little of the polyphonic structure of the music it encodes—and is therefore relatively inaccessible to non-specialists. Automatic polyphonic transcription into modern music notation can help unlock the corpus to a larger audience and thus facilitate musicological research.
“In this study we present four variants of a machine-learning model for voice separation and duration reconstruction in 16th-century lute tablature. These models are intended to form the heart of an interactive system for automatic polyphonic transcription that can assist users in making editions tailored to their own preferences. Additionally, such models can provide new methods for analysing different aspects of polyphonic structure.”
The full article lays out the researchers’ modelling approaches and the advantages of each. They report their best model returns accuracy rates of 80 to 90 percent, so for modelers, it might be worth the $39 to check out the full article. We just think it’s nice to see machine learning used for such a unique and culturally valuable project.
Cynthia Murrell, December 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
The Importance of Google AI
December 23, 2015
According to Business Insider, we’ve all been overlooking something crucial about Google. Writer Lucinda Shen reports, “Top Internet Analyst: There Is One Thing About Google that Everyone Is Missing.” Shen cites an observation by prominent equity analyst Carlos Kirjner. She writes:
“Kirjner, that thing [that everyone else is missing] is AI at Google. ’Nobody is paying attention to that because it is not an issue that will play out in the next few quarters, but longer term it is a big, big opportunity for them,’ he said. ‘Google’s investments in artificial intelligence, above and beyond the use of machine learning to improve character, photo, video and sound classification, could be so revolutionary and transformational to the point of raising ethical questions.’
“Even if investors and analysts haven’t been closely monitoring Google’s developments in AI, the internet giant is devoted to the project. During the company’s third-quarter earnings call, CEO Sundar Pichai told investors the company planned to integrate AI more deeply within its core business.”
Google must be confident in its AI if it is deploying it across all its products, as reported. Shen recalls that the company made waves back in November, when it released the open-source AI platform TensorFlow. Is Google’s AI research about to take the world by storm?
Cynthia Murrell, December 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Gibiru Compromised?
December 22, 2015
I assume, gentle reader, that you are aware of the anonymizing search system called Gibiru. Today (December 22, 2015) I received this notification when I attempted to run a query about Palantir on this search system:
The Kaspersky information link is a 404. I located no substantive information about this possible issue when I poked around online. I had in my files a link to https://anonymous-gibiru.com/ which did not trigger the malicious file warning.
Stephen E Arnold, December 22, 2015
Text Analytics Jargon: You Too Can Be an Expert
December 22, 2015
Want to earn extra money as a text analytics expert? Need to drop some cool terms like Latent Dirichlet Allocation at a holiday function? Navigate to “Text Analytics: 15 Terms You Should Know Surrounding ERP.” The article will make clear some essential terms. I am not sure the enterprise resource planning crowd will be up to speed on probabilistic latent semantic analysis, but the buzzword will definitely catch everyone’s attention. If you party in certain circles, you might end up with a consulting job at mid tier services firm or, better yet, land several million in venture funding to dance with Dirichlet.
Stephen E Arnold, December 22, 2015
-
- Subscribe to Beyond Search
Feature archive
News archive
-
