MarkLogic: A NoSQL Vertical Jump for More Revenue?
July 24, 2019
Is NoSQL-database-platform-firm MarkLogic is emulating Dialog Information Services and Lexis Nexis or vertical plays for quirky controversial, niche markets like drugs and medical device specific services? MarkLogic’s push into other verticals like professional publishing and finance have not generated the type of buzz and revenue that other Silicon Valley firms have sparked. Maybe pharma is the key which will unlock massive returns for the stakeholders? MarkLogic has resisted the type of acquisition and repositioning play that kCura executed in eDiscovery? Perhaps pharma, a sector whose revenue grows as the number of global players shrinks?
The company announced the MarkLogic Pharma Research Hub, created to bring the power of federated search to the field of pharmaceutical R&D. The product description tells us:
“For pharmaceutical companies, the discovery of new molecules and the cost of developing a successful medicine can take up to 15 years and $2.6 billion — slowing potentially life-saving drugs from getting to the patients who need them and resulting in abandonment of drug trials when faced with potential failure. In this industry, even small improvements to streamline R&D processes can lead to substantially higher revenue and lower costs. To achieve those goals, pharmaceutical companies need to leverage their massive data assets that include decades of research and clinical trial data. The challenge is that researchers are often unable to access the information they need. And, even when data does get consolidated, researchers find it difficult to sift through it all and make sense of it in order to confidently draw the right conclusions and share the right results.”
The product announcement elaborated:
“The main challenge facing IT departments that serve pharma R&D is patchwork infrastructure that creates the data silos that isolate and restrict access to data. Pharmas need to leverage massive data sets, including decades of research and clinical trials information.”
In addition, we’re reminded, disparate data silos hamper collaboration, upon which researchers rely heavily. The announcement goes on to outline the platform’s advanced features: the ability to load any pharmaceutical data set, relationship visualizations and discovery, and customizable search results. Naturally, these functions are made possible by machine-learning AI.
Founded in 2001 as Cerisent, MarkLogic is based in San Carlos, California, with several offices in the U.S. and in Europe. After changing its name, it released Version 1 of its platform in 2003. The company has ingested more than $170 million in venture funding. The firm has probed the intelligence sector and marketed itself as an enterprise search solution. But revenues? MarkLogic is a privately held firm just 18 years young.
Cynthia Murrell, July 23, 2019
Factualities, July 24, 2019
July 24, 2019
There were not too many facts for the DarkCyber team to extract from the Facebook Versa hearing. Bummer. We did find a few, including the craziest Factuality of the week.
$152 billion. Revenue from game streaming in 2019. Source: Fortune Magazine
Here are other numerical gems we spotted in the last week. Read on.
7. The number of stalkerware apps (downloaded 130,000 times by consumers) removed from the Google Play Store by Google itself. Source: 9to5 Google
10. Number of months Google’s blog management tool survived before termination with extreme prejudice. Source: Android Police
40. The increase in hate speech on 4chan in the last 12 months. Source: Vice
50. The percentage of time “older” Americans spend alone. (Note: No data about the presence of a mobile phone owned by this group.) Source: Pew Research
68. Percentage of information technology managers who cannot keep up with cyber attacks. Source: OodaLoop
1,000. Number of private conversations record via Google Assistant leaked. Source: CNBC
1,500. Number of digital textbooks Pearson (which once owned a wax museum) will be digital first. Source: BBC
3,200. Number of changes to “search” Google made in 2018. Source: Search Engine Roundtable
700,000. Number of podcasts available at this time. Source: New York Times
$1,800,000. Amount bad actors demand from a college compromised by ransomware. Source: Naked Security
$10,000,000. Amount a now-former Microsoft engineer stole via customer fraud. Source: The Register
13,000,000. Number of daily users of Microsoft’s Slack clone service. Source: Slashdot
63,000,000. Number of Amazon Prime members as of July 2019. Source: ZDNet
$301,000,000. Amount stolen per month via business email compromises. We find the “1” a nice touch. It communicates accuracy. Source: Bleeping Computer
1,000,000,000. Number of installs of Word on Android OS mobile phones. Source: MSPowerUser
$4,500,000,000. Size of Google Venture’s arm. Source: Business Insider
$14,300,000,000. Size of the consumer drone market in 2029. Note: This is one tenth the size of today’s game streaming market. Source: Reuters
$45,000,000,000. Cost of financial crime in 2018. Source: Dark Reading
We still admire the “one” in the estimate of losses from phishing. Precision is good, especially when a “one” is involved.
Stephen E Arnold, July 24, 2019
Smart Software: About Those Methods?
July 23, 2019
An interesting paper germane to machine learning and smart software is available from Arxiv.org. The title? “Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches”.
The punch line for this academic document is, in the view of DarkCyber:
No way.
Your view may be different, but you will have to read the document, check out the diagrams, and scan the supporting information available on Github at this link.
The main idea is:
In this work, we report the results of a systematic analysis of algorithmic proposals for top-n recommendation tasks. Specifically, we considered 18 algorithms that were presented at top-level research conferences in the last years. Only 7 of them could be reproduced with reasonable effort. For these methods, it however turned out that 6 of them can often be outperformed with comparably simple heuristic methods, e.g., based on nearest-neighbor or graph-based techniques. The remaining one clearly outperformed the baselines but did not consistently outperform a well-tuned non-neural linear ranking method. Overall, our work sheds light on a number of potential problems in today’s machine learning scholarship and calls for improved scientific practices in this area.
So back to my summary, “No way.”
Here’s a “oh, how interesting chart.” Note the spikes:
Several observations:
- In an effort to get something to work, those who think in terms of algorithms take shortcuts; that is, operate in a clever way to produce something that’s good enough. “Good enough” is pretty much a C grade or “passing.”
- Math whiz hand waving and MBA / lawyer ignorance of what human judgments operate within an algorithmic operation guarantee that “good enough” becomes “Let’s see if this makes money.” You can substitute “reduce costs” if you wish. No big difference.
- Users accept whatever outputs a smart system deliver. Most people believe that “computers are right.” There’s nothing DarkCyber can do to make people more aware.
- Algorithms can be fiddled in the following ways: [a] Let these numerical recipes and the idiosyncrasies of calculation will just do their thing; for example, drift off in a weird direction or produce the equivalent of white noise; [b] get skewed because of the data flowing into the system automagically (very risky) or via human subject matter experts (also very risky); [c] the programmers implementing the algorithm focus on the code, speed, and deadline, not how the outputs flow; for example, k-means can be really mean and Bayesian methods can bay at the moon.
Net net: Worth reading this analysis.
Stephen E Arnold, July 23, 2019
Sockpuppet Image Source
July 23, 2019
I read “Turn Selfies into Classical Portraits with the AI That Fuels Deepfakes.” I gave the system a spin. I uploaded a picture from this week’s DarkCyber. The system generated a wonderful image usable by anyone with access to a source of images; for example, Bing Images or Facebook. Here’s the result:
Working well. Cloud centric or a laptop? I loved the explanation: “Huge traffic.” Back to those scaling lectures.
Stephen E Arnold, July 23, 2019
Search Engine Optimization: Why Search Delivers Irrelevant Results and Ad Budgets Can Go Poof
July 23, 2019
DarkCyber noted “What Are Click Farms? A Shadowy Internet Industry Is Booming in China.” A diligent “real news” professional noted that one can buy clicks. DarkCyber spotted these services on gig economy sites like SEOExperts, Fiverr, and similar services some time ago. Think in terms of years.
The write up explains: Click farms
are plugged in and programmed to search, click, and download a certain app over and over again. The goal is to manipulate the system of app store rankings and search results.
The procedure is:
Click farms use an automated process hacks into the normal App Store Optimization (ASO) practice — which requires developers to use certain keywords in descriptions and attract users by being a useful product — and are programmed to promote apps by imitating a real user by searching for certain keywords, clicking on the app, downloading, and even writing positive reviews.
The write up focuses on apps and China.
DarkCyber wants to suggest that click farms are available to perform tasks like these:
- Target a company’s online ads, click on them, and burn through the budget for a keyword so a second place owner of a keyword pops up and presumably gets the “real” clicks from an actual interested person. (Keep in mind that a savvy competitor can have this technique used against his or her campaign.)
- Target a concept and click links. The result is what DarkCyber and its beloved leader calls “augmentext.” The idea is that a concept, not a site, can be converted into an attractor for a Google-type relevance system
- Click on an entity and cause that entity to have “magnetism.” With the loopholes and weaknesses inherent in the core algorithms, an entity can become “hot” or a “trend.”
The write up points out that click farms are illegal. Perhaps the estimable search engine optimization industry should police its behaviors? Perhaps online disinformation consultants should not use these services?
I am not sure that click farms are new, particularly shadowy, or going to go away. Spoofing relevance is too darned easy and there’s zero incentive for certain vendors selling ads or offering to manipulate opinion to change.
Stephen E Arnold, July 23, 2019
DarkCyber for July 23, 2019, Now Available
July 23, 2019
DarkCyber for July 23, 2019, is now available at www.arnoldit.com/wordpress and on Vimeo at https://www.vimeo.com/349282829. The program is a production of Stephen E Arnold. It is the only weekly video news shows focusing on the Dark Web, cybercrime, and lesser known Internet services.
This week’s DarkCyber reports about Australia’s use of its anti-encryption law; tools for video piracy, a profile of SearchLight Security’s Cerberus system, and where to get information needed to join a Dark Web forum.
This week’s lead story concern easily findable software to facilitate video piracy and streaming. A report in TorrentFreak presents information from an unnamed source. This individual allegedly has been involved in video piracy and streaming for an extended period of time. The individual provides specific information about some of the software needed to remove digital rights management protections from commercial, copyrighted video content. The DarkCyber research team was able to locate software designed for the same purpose. No Dark Web and Tor were required. More significantly, these programs can be located by anyone with access to a browser and a Web search engine like Bing, Google, or Yandex. DarkCyber’s research has revealed that industrialized crime is now playing a larger role in streaming stolen video content.
Other stories in the July 23, 2019, program are:
First, Australia’s anti encryption law is now being put to use. The new regulations were used in the warrant to obtain content from a journalist. Australia is a member of the Five Eyes confederation. Australia’s law requires companies to cooperate with law enforcement and provide access to encrypted and other secured information. Canada, New Zealand, the United Kingdom, and the United States are likely to have elected officials who will seek to implement similar laws. News organizations in Australia perceive such laws as a threat.
Second, DarkCyber profiles a company founded in 2017 focused on providing law enforcement and intelligence professionals with an investigative tool. The company indexes a range of content, including forums, Dark Web sites and services, and social media content. Plus the company has created an easy-to-use interface which allows an investigator or analyst to search for a person of interest, an entity, or an event. The system then generates outputs which are suitable for use in a legal matter. The company says that use of its system has grown rapidly, and that the Cerberus investigative system is one of the leaders in this software sector.
Finally, DarkCyber provides information about a new report from IntSights, a cyber-intelligence firm. The report includes information which helps an individual to gain access to “cracker” forums and discussion groups which examine topics such as credit card fraud, money laundering, contraband, and similar subjects. The video provides the information required to download this report.
DarkCyber videos appears each week through the September 30, 2019. A new series of videos will begin on November 1, 2019. Programs are available on Vimeo.com and YouTube.com.
A new series of DarkCyber begin in November 2019.
Kenny Toth, July 23, 2019
Amazonia, July 22, 2019
July 22, 2019
About that JEDI contract? The big news is that President Trump is going to check out the $10 billion deal for the Department of Defense’s cloud computing initiative. The driver of the Bezos bulldozer owns the Washington Post. Allegedly Mr. Trump refers to the prestigious “real news” outfit as “Amazon’s Washington Post.” A good sign? Who knows. Other Amazon items the DarkCyber team processed this week were less interesting. Here’s a few which seemed intriguing.
Amazon-SUE-ticals
Amazon wants patient data. (Note: With the patient data comes useful information about the prescription itself. Doctors in Florida, are you paying attention?) CNBC, which continues to surprise as a source of useful information, published “Amazon Threatens to Sue Major Pharmacy Player If It Prevents PillPack from Accessing Patient Drug Data.” We noted this statement in the write up:
PillPack was informed this week that it will soon be cut off from accessing that data via a third-party entity, ReMy Health — a move that could seriously complicate its business. Amazon is considering legal action against Surescripts to halt those efforts, said the people, who asked not to be identified because the deliberations are confidential. One person told CNBC that PillPack has already sent a cease-and-desist letter to Surescripts.
Several observations:
- Executives are fearful of Amazon. For a reason, read “Amazon Brand Control” below
- These data feed into other Amazon “areas of interest”. DarkCyber speculates that delivery information, compliance data, and policeware services may benefit
- Amazon doesn’t have a direct deal with Surescripts.
DarkCyber believes that Amazon’s “customers” may provide a bit of shadow power to make “sure” the information is provided. And if Surescripts decides to sue Amazon in an expensive, lengthy court battle? A deal may result. Worth monitoring this pharma-SUE-tical matter? Yep.
The Pesky EU and Amazon
“Antitrust: Commission Opens Investigation into Possible Anti-Competitive Conduct of Amazon” makes clear that the European Commission has a new project for some of its lawyers, INSEAD graduates, and accountants: Amazon. Here’s the problem:
Amazon has a dual role as a platform: (i) it sells products on its website as a retailer; and (ii) it provides a marketplace where independent sellers can sell products directly to consumers.
Is Amazon a monopoly? Judge for yourself by reading “Amazon Brand Control.”
Amazon Brand Control
The Rupert Murdoch “real news” outfit published “Amazon Seeks More Brand Control.” DarkCyber thought the story left an important point unstated; for example, monopolies exercise their power directly and by fiat. The “real news” outfit reported:
The program — which allows brand rights to be bought for a fixed price on 60 days’; notice—… is part of a push by Amazon to obtain a stable of exclusive brands for the platform.
What happens if a “brand” does not want to play ball? Well, there’s eBay, driving for Uber, or an Amazon warehouse job. You can read the write up for free if you can find the dead tree version of the Murdoch property for Friday, July 19, 2019, B-1. If not, you can click here but you may have to pay. “Cutting out the middleman” is a nice way of saying, “My way or the highway.” A rose by any other name is still a — Prime day rose?
Bloomberg Identifies Amazon’s Most Serious Research Project
Bloomberg’s judgment can be measured against its reports of spy chips on motherboards. Now the company has turned its attention to Amazon’s research projects. Forget the policeware and intelware activities. The rubber hits the door mat with Amazon’s retail store experiments. You can get the Bloomberg analysis of Amazon’s “most ambitious research project” in this July 18, 2019, essay/analysis. Note that you may have to pay for this insight. The write up states:
Will all this work be worth it? Some Go stores seem almost deserted except for the lunchtime rush. Employees familiar with Amazon’s internal projections say the outlets in Chicago, in particular, are falling short of expectations, and the company has had to resort to raffles and giveaways of tote bags and other branded goodies. Yet, as the turbulent history of the project suggests, the Go store isn’t so much the culmination of the company’s efforts but something closer to an ongoing experiment.
Plus, there’s a picture:
DarkCyber heard that in one Go store, humans were added because theft was an issue.
Amazon Police Map
“Here’s The Most Complete Map So Far Of Amazon’s Ring Camera Surveillance Partnerships With Local Police” looks like this:
Is Amazon in the policeware business? You judge for yourself by checking out this mostly ignored item. Also, how many of these “installations” are trials, freebies, and demonstrations? Some trial are ending; for example, Orlando’s.
For more on this topic, DarkCyber offers a for fee webinar on this topic. Write us at darkcyber333 at yandex dot com.
Prime Day Data
DarkCyber has no way of knowing if the data in “Amazon Just Announced Prime Day Data, and the Staggering Numbers Beat Black Friday and Cyber Monday Combined” are accurate. But the numbers do seem to be beefier than those reported by Nordstrom and other outfits of that ilk.
So how big? Well…
- 175 million items sold
- 175 million Prime members
- Each Prime member bought 1.75 items
Do these numbers look similar? Sophisticated analysis for sure.
First Transnational Bank of Amazon
The FTBA does not exist yet, but some think it may arrive. “Can Apple, Google, Facebook and Amazon Transform Banking? Yes, and They’re Closer Than You Think” states:
Amazon’s competitive advantage is its ability to build cloud-banking much more securely than banks. It’s leading in the cloud, so this means your banking would no longer need to be local, it can be global. One account for all currencies.
The write up even suggests that one obtain a consultant’s research report to make the case for FTBA. Objective? Sure. DarkCyber believes everything its team reads on the Internet, including ITPortal’s analyses.
Amazon Fee Triggers
Amazon published in April 2019 a paper called “AWS Reliability Primer.” The idea is that one must consider how much of each of these “values” an AWS developer requires:
- Operational excellence
- Security
- Reliability
- Performance efficiency
- Cost optimization.
From a technical or architectural point of view, the write up provides useful information about the linkage between what Amazon can deliver and what one’s budget can tolerate.
DarkCyber thinks that this list of five factors explained in 62 pages of text highlights where costs can skyrocket if the AWS “customer” makes bad decisions.
Best practices or we warned you? You decide.
Amazon Stock Value
Seeking Alpha knows that fear, uncertainty, and doubt are good for some businesses. “Amazon’s Slowing Growth May Sink The Stock Following Results” opines:
AMZN is seeing a deceleration of growth in many of its business units. It could result in the stock pulling back following the results to around $1,800 based on an analysis of the chart.
Disaster looms, but one can tap Seeking Alpha for financial advice.
Amazon Satellites
Is this a $100 billion per year business? Motley Fool (“fool” in shorthand) states:
Amazon confirmed its plans, saying, “Project Kuiper is a new initiative to launch a constellation of low Earth orbit satellites that will provide low-latency, high-speed broadband connectivity to unserved and underserved communities around the world,” according to an Amazon spokesperson.
Satellites may be more reliable than floating Loon balloons.
Amazon Sales Reorg
Seeking Alpha published “AWS Reorganizes Sales Leader.” The site reported:
Web Services shuffled its sales team’s senior leadership earlier this year to clarify roles and eliminate confusion of multiple pitches to the same customer.
Criticism of AWS Firecracker?
Tech Republic’s “The Clearest Sign of AWS’ Open Source Success Wasn’t Built by Amazon” seems to be critical of the Bezos bulldozer. The write up states:
AWS Firecracker is great open source technology, but the best indication of its open source success is what Weaveworks built on top of it.
We think this means that Amazon provided a foundation, and another company used that foundation to create a successful solution.
The write up is a bit convoluted, and it preserves Tech Republic’s ability to keep the doors open to content sponsors.
Partners and Integrators
Datadog is now competent in AWS migration. Source: MarketWatch
Northern Virginia Community College will train US Marines to use AWS. Amazon seems to have some confidence in its winning the JEDI competition. Or, this could just be another “train people to use Amazon” play. Source: World Socialist Web Site (real news all the time we assume)
SnapLogic offers a quick start for those wanting to put a data lake on AWS. It appears that SnapLogic will work with Agilisium. Source: Help Net Security
ZenDesk moves to make customer data more actionable. We are not sure what “actionable” means, but with an expanded AWS service, DarkCyber has high hopes for understanding the concept. Source: Yahoo
Stephen E Arnold, July 22, 2019
Google in Space: A Sling, Not a Loon Balloon
July 22, 2019
I read “Alphabet-Backed Space Launch Company Wins Pentagon Contract.” The write up explained that Google, Airbus, and a venture capital company are backing a company called Spinlaunch. Good name for an outfit which proposes to throw small satellites into orbit. None of the messy rocket fuel needed upon which are fixated Messrs. Bezos and Musk. Spinning is allegedly better than blasting.
Here’s the approach:
SpinLaunch emphasizes a “ground-based kinetic energy” approach. Specifically, it proposes to use a centrifuge-like device to spin a rocket round and round like a slingshot, building up momentum, and then hurtling it into the sky at hypersonic speed.
I assume the Loon balloons cannot get sufficient altitude to provide a platform from which to place tiny satellites in orbit. Hence, a new approach which seems similar to David’s use of a sling and stone to slay Goliath. Or, in the case of the Bezos and Musk “Goliaths” a centrifuge repurposed. Worth watching out for. Slings can be tough to control and David’s toss at the metaphorical Goliath may have been due to chance, not satellite worthy accuracy. Throwing a tiny satellite into the space junk orbiting the earth might score a chance hit. Interesting to consider how a satellite owner would react to a Spinlaunch stone knocking out a very expensive satellite.
Stephen E Arnold, July 22, 2019
Google and Privacy: Quote to Note
July 21, 2019
Google and privacy. Sounds good. Like ham and cheese, bacon and eggs, cyber and space. “Google Is Stopping Websites From Tracking Users in Incognito” revealed “Google said it will stop websites from being able to identify whether you’re browsing using its private incognito mode in a forthcoming browser update.”
Allegedly Google said:
‘People choose to browse the web privately for many reasons. Some wish to protect their privacy on shared or borrowed devices, or to exclude certain activities from their browsing histories,’ said Google in a statement. ‘We want you to be able to access the web privately, with the assurance that your choice to do so is private as well.’
Why wouldn’t everyone believe Google after its brilliant, unflustered testimony in which the company’s spokes human refused to commit to a third party audit. See “Google Executive Refuses to Commit to Independent Audit During Senate Hearing.”
Are you a doubter? DarkCyber believes everything it reads on the Internet, including Facebook, Google, and IBM content. Google just revealed that it is going to achieve quantum supremacy this year.
Oh, one minor question: Does that apply to Google’s tracking of its users?
Believe it.
Stephen E Arnold, July 21, 2019
Facebook: Fighting the Good Ad Fight
July 21, 2019
It is search to the rescue! Following a settlement meant to eliminate discrimination on Facebook last year, the company is amending how it delivers housing, job, and financial services ads. ABC News reports, “Facebook to Make Jobs, Credit Ads Searchable for US Users.” The platform makes most of its money from targeted advertising, but the technique has its problems. Reporter Frank Bajak writes:
“The move is likely part of Facebook’s strategy to show regulators that is doing a good job policing its own service — putting it in compliance with existing anti-discrimination law — and doesn’t need a heavy-handed approach from lawmakers. It comes as the company is facing increasing regulatory pressures.
As part of the settlement with plaintiffs including the ACLU and the National Fair Housing Alliance, Facebook agreed in March to stop targeting people based on age, gender and zip code and to also eliminate such categories as national origin and sexual orientation. The groups had sued claiming Facebook violated anti-discrimination laws by preventing audiences including single mothers and the disabled from seeing many housing ads — while some job ads were not reaching women and older workers. Galen Sherwin, senior staff attorney at the ACLU and the group’s lead attorney in the case, said making the three Facebook databases searchable by anyone ‘definitely creates greater access to information about economic opportunities.’”
Of course, there could still be a lot of bias hidden in those ad-steering algorithms, but good luck achieving complete transparency there—proprietary software and all that. Besides, there are also the issues of privacy, anti-trust violations, and hate speech to consider. At least Facebook appears to be looking ahead: they say they are fighting voter suppression efforts and potential attempts to interfere with the 2020 census. Will it be enough to keep its critics, like the ACLU and the National Fair Housing Alliance, at bay?
Cynthia Murrell, July 21, 2019
-
- Subscribe to Beyond Search
Feature archive
News archive
-
