Berkeley-based Wise.io has launched its first product for data scientists and developers as part of a goal to make machine learning technology more accessible.

The idea for the initial “machine learning as a service” product came from the founders’ experiences on Kaggle, the website that 70,000 data scientists use to compete to solve tricky math problems.

CEO Joshua Bloom noticed that data scientists were using a specific algorithm on Kaggle to win competitions, but it was clunky to use. “We wanted to make it as fast as possible and in-memory efficient,” Bloom told me. The Wise.io team [above] have some impressive credentials: They have Ph.Ds in statistics, computer science, and astrophysics.

An added benefit: Many of the data scientists that are using Kaggle have day jobs at large companies. To reach enterprise customers, Bloom is hoping that these developers will push their IT departments to buy a site license.

“For now, the lowest barrier to entry are those who are using machine learning algorithms in their daily lives,” said Damian Eads, Wise.io’s cofounder and engineering lead. Eads would not disclose the existing customers of the new product, but he said the company is already profitable, a rare feat in Silicon Valley.

The company also plans to make money by offering professional consulting services to large companies who are looking to invest in a machine learning implementation.

Wise.io doesn’t have any immediate plans to raise institutional funding, and is an entirely bootstrapped effort.

---

Big Data and Predictive/Real-time Analytics startups: Are you looking to jumpstart development & accelerate market traction? Sign up for the SAP Startup Focus program to receive technology, support, resources and community to help you develop new applications on SAP HANA, a cutting edge database platform. Get started here, and enter promo code “VB2013″ on the form.

---

Facebook’s presentation — and images — were curiously void of detail on how ads will fit into the new Facebook, but you can bet it’s been a hugely significant part of the conversation as Facebook continues to seek lucrative but appropriate ways of monetizing the social network.

A Facebook representative did tell me via email that businesses now have an “even more visually rich way to showcase content” and that users respond better to more visual stories. In addition, she said, your cover photo is going to be even more important in the new reality, as it will show up in the news feed in all its glory — or lack thereof — when a friend likes a business:

One superimportant point: The new news feed is NOT being accompanied by a new algorithm.

In other words, Facebook is not using this roll-out as an opportunity to revamp EdgeRank, the programmatic method by which Facebook determines how interesting you are, how interesting your content is, and therefore and most critically, how often and where to display your messages.

The early reports from advertisers are agreement with Facebook’s messaging.

“Facebook’s news is unbelievably exciting,” Fiksu CEO Micah Adler says. “Introducing enhanced feeds creates rich contextual advertising opportunities and serves as a perfect complement to Facebook’s unprecedented targeting and measurement capabilities.”

Nothing matters more to advertisers than context, Adler says, saying that while Facebook likes are simple, one-dimensional indications of positive sentiment, interactions between advertisers and people through contextual targeting in the new news feeds will “inevitably increase buyer interest.” He added that Fikus is planning to invest “significantly in new ad units on Facebook, particularly in mobile.

“The new look for Facebook’s News Feed promises new opportunities for advertisers,” Adobe media and ads director Justin Merickel said. “As users tend to show a stronger interest in active, personalized news feeds, embedded ads will likely be more relevant and generate higher ROI for advertisers.”

One concern? The feed choice options that Facebook is implementing.

“The friends-only feed could get a lot of the volume which may be a concern — but it does stop people getting upset with sponsored content appearing around their friends content,” Simon Mansell, the CEO of TBGdigital, said. “Standout will be harder for brands as a result but it’s probably a better user experience.”

Source: Facebook

Facebook’s new news feed options

That said, he added that “the chronological view is interesting — as it removes the argument that there is an algorithm in between users and content.”

Sarah Carter, the GM of social business at Actiance, echoed the concern, but she noted that overall, brands that truly get social marketing could do better with the new feed:

“People can skip content from a business if they want to,” Carter said. “The flip side is that the new design is meant to get people browsing for longer, and improve the overall Facebook experience. So if a business is able to effectively market to its audience on this channel, using content that is visual, relevant and spoken in an authentic voice, the ROI should be greater.”

One other tip from Facebook echoed Mansell’s comments: People will be able to choose the “Following” feed, which will simply display everything that every user and company they are following is publishing. That “reality-bubble free” zone in Facebook could get very interesting as people discover an unfiltered lens into their full social graph.

The new feed roll-out will happen “very slowly” over the coming weeks on mobile and web, Facebook says.

This is a guest post by Jon Soberg, managing director at Blumberg Capital.

Venture capitalists picking up the Moneyball strategy. It’s a fascinating idea, and I know plenty of my peers are working on algorithmic approaches to better their craft and gain an edge. As I see innumerable pitches that leverage “big data” as a key aspect of a business, and read about all of the ways that big data is changing the world, it seems natural that it will infiltrate the world of venture capital. Whether it works or not remains to be seen.

As someone who has worked at both ends of the spectrum — as a CFA performing public equity analysis, and an early stage investor — this is more than just a passing interest for me. I pay close attention to data, and I firmly believe that people who understand the numbers better than their competitors, will ultimately win. In public equities, some groups have clearly found ways to outperform using algorithms over time periods.

Are algorithms the future of VC?

In VC, we know the stats; a large percentage of companies fail, and returns have lagged over the past several years (although some firms have continued to outperform). Is data analytics the answer as some have suggested? Is it key to success in our industry, which is clearly changing? I have to say no, but will qualify my answer.

At the early stage where I invest, I am often assessing a team with a product that is not yet in the market, with no revenue, and with plans to hire some significant team members within the first year. Small data. There are very, let me stress VERY, few data points. And most of the data points that are there are qualitative, and there is plenty of noise. It hasn’t stopped me from building a few models along the way to see if I can glean some insight, but I can tell you what the models yield — garbage in, garbage out.

One might argue that algorithms can help predict markets, and I believe that could be true, but the size of the market opportunity for most companies is defined not by the actual size of the market, but by the team’s ability to address the market. It is SAM, not TAM. Let’s say a model detects that a market is developing, can the model help select the team that will win in that market? Revenue and growth are tightly coupled with a team’s ability to understand the needs of their customers and to fulfill those needs. It comes down to the team. I have yet to see a model that can tell me that a Mike Lazerow, Jason Goldberg, or Joe Lonsdale are better at figuring out their markets than others, but I can tell you that if you meet with any of them in person, you will know.

Early-stage vs. later-stage investing

That said, I do think they can be helpful in some ways, especially in later stage investments. One of my toughest challenges is assessing opportunity cost. If an algorithm can help me decide which of several investments has the best potential, then it can be valuable. But, in my experience, the value is directly linked to the maturity of the company, as small data becomes bigger data. And the breadth of the comparison set is critical to get a good signal; any bad data in these models can really skew results.

One algorithmic approach that I believe has potential is employed by Correlation Ventures. They analyze investments made by other VC’s using their proprietary algorithm, and join some of the rounds. Just the fact that they are leveraging the work done by other VC’s gives them an inherent advantage in selection — they add another layer to existing analysis.

Data won’t help us find the next Zuckerberg

The combination of human and algorithmic analysis is the best way to use data analytics. But data analytics won’t help us score the home run hits that are absolutely essential. The Moneyball approach simply doesn’t work because hit rates are not high enough to manufacture returns by getting on base. The key is swinging hard every time you step up to the plate — a Facebook home run is like hitting the ball OVER McCovey Cove and getting 10 runs for it.

So when I look at how data analytics will impact VC, do I realistically think algorithms will give me a competitive advantage? No. These are not liquid markets, and there are very few good data points. Will it help with due diligence and adding a few data points for my analysis? Maybe. Will it help identify target markets? Potentially. Will it make me a better investor? Doubtful. Will it help me win deals versus our competitors? No. Will it help raise money? To be determined.

To quote one of my mentors, Wharton Faculty member Dave Pottruck, “Go with your gut, but always crunch the numbers first.” But in the absence of all that much valuable data, VC’s need to rely on old school-hustle, homework, and instinct.

Jon Soberg is a Managing Director at Blumberg Capital, where he invests in early stage companies, specializing in FinTech, SaaS, and eCommerce. Prior to joining Blumberg Capital, Jon has been a serial entrepreneur and senior executive in multiple companies including Ditech, Broadband Digital Group and Adforce, which had a highly successful IPO.  

A CFA Charterholder and adjunct faculty in the Wharton Marketing Department, Jon earned a B.S in Engineering from Harvey Mudd College, an M.S. in Engineering from Northwestern University, and an MBA in Entrepreneurial Management and Marketing from the Wharton School, where he is a Palmer Scholar.

Top Image via Shutterstock

---

Big Data and Predictive/Real-time Analytics startups: Are you looking to jumpstart development & accelerate market traction? Sign up for the SAP Startup Focus program to receive technology, support, resources and community to help you develop new applications on SAP HANA, a cutting edge database platform. Get started here, and enter promo code “VB2013″ on the form.

---

Is your startup struggling to solve a tricky data challenge? Data science company Kaggle is launching a program today to rally its community of 70,000 statisticians from around the world to solve your most enduring problems.

The only catch? The new program known as “let the crowd be your cofounder” is only open to small startups, and founders will need to set aside $5,000 in prize money or 0.5 percent in company equity for the winning submission. Still, it’s a darn sight cheaper than hiring a full data services team.

Kaggle is a San Francisco-based startup that hosts data science competitions, typically for large companies and nonprofits like the Hewlett Foundation, AllState, and GE. For instance, The Hewlett Foundation offered up $100,000 in prize money to the first person that could create an algorithm for scoring essays the same way a human grader would.

Related: Read more about cool Kaggle competitions here.

Kaggle’s CEO Anthony Goldbloom told me he realized that fast-growing tech startups are storing vast volumes of consumer data but often don’t know what to do with it.

“I got really excited about the idea of data-driven startup just as I was starting Kaggle,” said Goldbloom in an interview. Before launching the program to the public, Goldbloom worked closely with three startups to solve their data challenges:

Researchers led by Krishna Shenoy, a professor of electrical engineering, bioengineering, and neurobiology at Stanford, doubled the efficiency of implanted neural prosthetics to allow brains to control external systems, simply via thought.

In other words, moving things with your mind.

But it’s no Jedi mind trick, and it doesn’t even involve humans at this point, although human brain sensors that drive robotic limbs have been built. Rather, the experiments involved rhesus monkeys with sensors embedded in their skulls.

What makes it possible is that in both monkey and humans, neurons fire in recognizable patterns when we want to move our hands, feet, a robotic arm, or a cursor on a computer screen. By intercepting and decoding those signals, scientists have been able to allow people to control computerized systems simply by thinking.

The primary use case, of course, is people who are paralyzed or who have lost a limb. Eventually, researchers believe they will be able to give disabled veterans or paraplegics the ability to control powered electronic prostheses … hopefully returning their ability to grasp, move, walk, and control their world.

Imagine an Oscar Pistorius, in the next summer Olympics, with smart legs and feet that move naturally, controlled by his brain in exactly the same way as his hands and eyes always have been.

The massive improvement in Shenoy’s work, according to Science Daily, is due to a new algorithm, called ReFIT, that does a better job of interpreting neural signals — twice as good as previous attempts — and a focus on groups on neurons that fire when, for instance, the brain is attempting to signal a limb to move, rather than just single neurons.

ReFIT achieves 75-85 percent of the speed of actual arms, which makes “neuroprosthetic” devices increasingly viable, and learns over time from users’ corrections to get better, quicker.

Which is great news for those who need prosthetics in the future, and one step closer to mind-controlled robots.

photo credit: DevanJedi via photopin cc

A Moneyball-style revolution is taking place in venture capital.

Just as the renegade general manager of the Oakland A’s flouted assumptions about baseball and replaced gut feelings and outdated statistics with more effective quantitative analysis, a new breed of venture capital firms are throwing out their Magic 8-balls and are using computer-based models to make smarter investments.

“The bottom line is that the game has changed … and there is a lot of digital exhaust out there,” said Chris Farmer, a partner at General Catalyst, which is considered one of the more progressive of the older VC firms.

Before it pours thousands of dollars into researching a potential investment (work typically performed by a well-paid associate), a small cadre of venture firms are using analytics tools to pull in megabytes of relevant data, whether it’s a game’s performance in the various mobile app stores or conversations about a new e-commerce site on Facebook and Twitter.

Relying on instinct simply isn’t good enough anymore.

If algorithms can predict the results of elections, why not the success or failure of a tech startup? Washington D.C. woke up to the power of data when numbers-cruncher Nate Silver proved critics wrong and delivered a gut punch to traditional punditry by accurately predicting the outcome of the presidential election.

In Silicon Valley, new firms are going a step further by creating an entire investment thesis around data. “Algorithms will be the heart and soul of due diligence — it’s not just a sanity-check mechanism,” said Matt Oguz, the managing partner of new investment firm Palo Alto Venture Science. “It’s the only way to cut through human bias.”

Oguz is taking a cue from Wall Street, which has been using algorithms for years to track the rise and fall of stock, and the macro-shifts in the financial markets. According to him, it’s only a matter of time before this algorithmic approach seeps into the private investment market like osmosis.

He is one of a growing number of investors developing algorithms. Some are vocal about their research, while others are keeping it quiet to prevent competing firms from following suit.

This next generation of venture capitalists are fixated on a billion-dollar question: Can an algorithm predict whether a tech startup will succeed or fail?

Venture capital is a numbers game

I recently received an invitation to meet a partner of a leading venture capital firms at a San Francisco coffee shop. The investor popped open his laptop to reveal a snippet of a “stealth” project he’s been working on for years.

At face value, it did not appear to me much more than a series of nondescript charts and graphs. However, he explained that this is the nascent research behind an algorithm that can take much of the guesswork out of venture capital.

In response to my befuddled gaze, he traced his finger over a graph that charts the rapid ascent of Facebook. At a certain point (marked in red on the graph), the algorithm triggers an alert: Facebook has become its own market. At that point, it is wise to invest in a company that would make money by piggy-backing off the social network, like Buddy Media. This is the most basic approximation of how the data might work to his advantage — but it illustrates the point.

This investor requested to remain anonymous. He is in no rush to pass on his research to competing firms with budget to throw at the problem.

Pages: 1 2 3

This is a guest post by social data expert Nick Arnett

Think Washington power brokers are ready to pay big bucks for Nate Silver’s predictions? Probably not – his numbers are too good. Yes, too good.

A decade ago, as I was starting one of the first social media analytics firms hoping to sell to politicos, a friend in Congress pointed out that even though politicians have to go outside to see which way the wind is blowing before making a decision, they ultimately only want polls that reinforce their beliefs.

Nancy Pelosi gave a fine example on The Daily Show when she dismissed Silver’s prediction that Republicans would still control the House. “That’s why we have elections,” she said.

Organizations rarely look to outsiders – and too often, even insiders – for accurate predictions, unless it is to shift blame. To the side that’s going to lose (and they all do, eventually), Silver is like a consultant hired to justify a huge investment in a horrible product, after the fact. The last thing the client wants is an accurate forecast.

The media doesn’t want accurate predictions either — consensus is boring. An accurate prediction is only newsworthy if there’s another poll that disagrees. Even more to the point, conflicting polls support conflicting people, whose arguments are the staple of news talk.

As many have already observed about Silver, the real sin of being repeatedly right is that his predictions rob us of the sense that we have control. Although helplessness is a big source of stress for anyone, it is huge for politicians and other leaders completely incompatible with success. For example, can anybody imagine that Steve Jobs would care what a poll predicted?

Quite the opposite – Jobs hired Trip Hawkins as Apple’s first head of marketing after Hawkins sent him a forecast doubting the size of the market. Jobs rewarded those who challenged his thinking; he had no interest in accurate predictions unless they came from creative thinking. Even then, it was the creativity that mattered. (Until I figured this out about him, I couldn’t understand why Jobs continued to tolerate, even encourage, my own pointed questions and challenges).

Still, Silver’s exceptional results truly matter.  He gives us a much-needed demonstration that the average “rock star data scientist” is just that — average. In a world where what we see, eat and otherwise use or are used by is driven by statistical analyses, our leaders’ hubris is an ever-present temptation for analysts to follow.

The stakes can be very high. Believe me, I know. A company affiliated with the Department of Homeland Security (DHS) bought my startup’s patents and I was told that DHS applied them to “terrorist buzz.” Two years later, Nov. 10, 2004, my niece’s husband, a Marine, was killed in action in Iraq. Since then, I’ve had to wonder if my own inventions were misused to justify the war that killed him.

Keep it up, Nate Silver. We need you because accuracy isn’t always profitable.

Nick Arnett is the Director of Product Management at NetBase. He was a participant in Carnegie Mellon Silicon Valley’s Disaster Management Initiative, working on citizen preparedness topics, especially crowd-sourced crisis mapping and vulnerable populations.

Prior to joining Netbase, he was the founder of Senti-Metrics Partners, a social media monitoring tool acquired by LiveWorld Inc.. 

Follow him on Twitter @NickArnett 

The world’s best number crunchers did more than predict the outcome of the presidential election. They may have helped a candidate get elected.

As Time’s Swampland reported, political staffer Jim Messina implemented a highly deliberate, number-driven campaign strategy. With Messina at the helm, Obama’s staff-members made decisions based on thorough quantitative analysis over experience or gut instinct.

For instance, with Hollywood showing overwhelming support for Obama, the campaign relied on math to determine which celebrities should host dinners. Data scientists noticed that George Clooney was idolized by women in their forties; this was the demographic group most likely to hand over cash in exchange for a dinner with Clooney and Obama. Sarah Jessica Parker, another dinner host, was chosen for her East Coast appeal and popularity with wealthy female liberals.

With the sheer volume of relevant data at our fingertips (tweets, news articles, demographic records, and so on) a pundit’s political instinct can only go so far.

“A few years ago the political parties — especially the Democrats — realized that algorithms could be used in politics,” said data scientist Jeremy Howard. Prior to the election, the Obama campaign issued a call for data scientists and turned up at universities like Stanford to recruit talent.

“They now identify areas of potential swing voters using algorithms, and they also figure out the optimal messaging for different kinds of voter using data analysis. Some of the country’s top analytic talent has been tapped by the parties to assist with this work,” he continued. Howard is the cofounder of Kaggle, a San Francisco-based startup that enlists the help of statisticians to solve tricky data problems.

This point was driven home when New York Times contributor Nate Silver was able to accurately predict the outcome of the election. His algorithm correctly projected the presidential vote in at least 49 states before a single ballot was cast. Silver put the probability of an Obama re-election at 90.9 percent.

Mitt Romney may not have prevailed in his bid to become the next president, but he did succeed in raising over $100 million in the first half of October alone. The Associated Press reported that he solicited the services of an analytics firm known as Buxton to identify rich and untapped potential donors. This was confirmed by the Fort Worth, Texas-based company’s CEO, Tom Buxton.

“This election will be remembered for the role that big data played behind the scenes, at every step. Obama won by identifying potential supporters and driving voter turnout in swing states,” said Tarun Wadhwa, a research fellow at Singularity University studying digital identification systems and their impact on governance.

“Future elections will build off this success, spread these tactics to other types of local and state races, and accumulate even more types of available information to process,” he said.

By storing and collecting data, the aids were able to put canny insights like these to use. Meet the data brigade that stole the show during this year’s presidential election:

[Image: Flickr user Joi]

July 9-10, 2013
San Francisco, CA

Early Bird Tickets on Sale

At 12 years old, Nick D’Aloisio began his career as a mobile programmer when he downloaded the Apple Developer Kit. By 15, one of his hacks, an app known as “Trimit,” caught the eye of an investor in Asia offering to fund “the company’s” next project.

What company? D’Aloisio, a school boy from South London, admitted to this venture capitalist that it was merely an after-school creation.

A whizz at the programming language C, D’Aloisio had created the app to test his skills in machine-learning technology, a branch of artificial intelligence that yields trends and patterns in a mass of data, and makes smarter predictions over time.

Summly’s founder demoes the app in San Francisco

Less than a year later, the “Internet boy wonder,” as he was dubbed by the European tech media after an appearance at Le Web’s conference, had developed a keen interest in web summarization technology. It’s a simple enough problem, but the execution is tricky. How do we take a meaty piece of content and whittle it down to its bare bones?

Today, D’Aloisio is ready to take his iPhone app public. Summly, available to download for free, reduces full-blown articles into snippets, making it easier to skim the news on a mobile device.

From ESPN to VentureBeat, the app includes almost all of the major news outlets. You lightly tap the screen to browse news, and if a summarized story piques your interest, you can share it with your friends via SMS, email, Twitter, or Facebook, or click a link the bottom of the screen to access the original, full-length story.

The founder told me he hit on the idea for the novel way to scan the news while studying for his History “GCSE” (an exam that all British school kids sit before their junior year). He quickly became frustrated by thousands of web search results that were slow to load and sometimes impossible to access on mobile.

“There has to be a way to summarize this information into bite-size,” he said. Trimit, the app he developed, summarized content into 140, 500 or 1,000 character summaries. With a little finessing, he moved on to Summly, which can automatically break down full pages of text into bullet-points. 

Most summarization experts have yet to apply their knowledge to a mainstream market. In a highly strategic move, D’Aloisio is applying the technology to the news, since, as he said, “No one had figured out how to perfect the news-reading experience on mobile.” In many cases, it’s relatively simple: Most journalists still write in the Associated Press (AP) style, so the first paragraph contains the crux of the news.

It’s not nearly as straight-forward as cropping the first few sentences, however. In the background, the technology works by scraping content, pulling out the most important bits, categorizing each story by topic, and displaying the news in a highly visual way. The summarized stories fit comfortably on your screen (tailored to both the iPhone 4 and 5).

At a coffee shop in San Francisco, D’Aloisio led me through a demo of the app: It’s a highly intuitive with the algorithm working to recommend news stories to you based on your pre-existing set of interests. As you shift location (from the UK to the U.S., for instance), the stories will change as well.

In person, D’Aloisio is wise beyond his years. It’s hard to believe this media-savvy, sharp-talking business executive is a teenager taking a leave of absence from school. He is both a high-school student and an employer. No longer a solo coder, he now has a team of seven people working for him. In addition, half a dozen researchers at SRI, the non-profit technology research institute based in Silicon Valley, are working with D’Aloisio to perfect the algorithm.

Some major technical hurdles  remain: Magazine-style features are tougher to summarize, and are often thrown out by the algorithm. The team is currently working on auto-translation (notoriously problematic and obvious to anyone who has ever used Google Translate, BabelFish and the like), and they have plans to integrate French and Spanish into the app. If they get it right, it could be a neat language-learning tool.

The team will also butt heads with Pulse, Flipboard, and a host of other beautifully-designed news-reading apps, especially since it is in the process of designing an interface for tablet devices. To succeed, it will need to win over voracious news consumers who routinely use web-based aggregators like Google Reader. The competition for these users is heating up: Circa is experimenting with ways to re-frame the news on mobile, and NewsBlur, a Y Combinator alum, recently launched its iPad app.

Summly is backed by several investors, including Horizons Ventures, Ashton Kutcher, Betaworks, Brian Chesky, Hosain Rahman, Joanna Shields, Josh Kushner, Mark Pincus, Matt Mullenweg, Stephen Fry, Troy Carter, Yoko Ono and many more.

Top image: Wired UK

Dabble in online dating? It turns out that the crippling fear of an awkward first date is the least of your troubles. A fraud is sweeping online dating sites, according to a special report in this month’s issue of Glamour Magazine.

The scam typically works like this: A con artist, usually based in an Internet cafe overseas, will lift a photo from Facebook or another social networking site. They will painstakingly craft a fake profile and begin targeting people that are looking for love.

Once they’ve made contact, they will typically request to move the conversation to a private instant messaging service. He or she will begin the courtship process by sending letters and love poems for a period of weeks and finally offer to fly to meet their victim. Within hours of the expected arrival time, an emergency will strike: A work visa has expired, or their aunt/niece/child is sick and they need a few thousand dollars to be wired over so they can finally meet their intended.

In many cases, scammers will choose to use pictures of military personnel. After discovering that his headshot consistently showed in hoax dating profiles (thanks to a Google alert), Army Master Sgt. C.J. Grisham set up a personal blog for soldiers to report their photo being used on online dating sites.

“Over the past few years, I’ve seen these scammers use all kinds of photos removed from open Facebook pages, blogs, official military websites, and command pages,” he wrote in a blog post last month. “I’ve also seen my own photos and name used.” (The image of Grisham that was used by scammers is pictured, left)

With a few of the largest player like OKCupid, Match, and others, there are precautionary measures in place. However, on smaller or niche dating sites like christiandatingforfree.com and datanta, there is no security system to spot a fraudster.

[Update: the site christiandatingforfree.com contacted us to notify us of the precautions they have in place, including a human editor who responds to complaints and reports, and ThreatMetrix, a cybercrime prevention software.]

“On some dating sites, as many as one out of 10 profiles is a scammer,,” Mark Brooks, editor of Online Personals Watch told Glamour.

The lesson here is that online dating startups will need to step up their game to keep consumers safe. “In the war against online dating scams and security threats, we’ve chosen to do whatever is necessary to always be a few steps ahead of scammers, and not the other way around – which is usually too late for our users,” said Cupid.com’s CEO, Bill Dobbie.

“This decision has its high price tag, but the alternative can be much costlier,” he continued. For this reason, Cupid.com, the online dating network behind Flirt.com and BlackMatch.com, announced a partnership this month with RecSys Ltd, an anti-scammer technology. The goal is to catch them 35 percent faster.

With the new system, any individual identified as a scammer is added to the master database, which other online dating companies can access. To identify rogue behavior, the algorithm factors in the user’s login location, IP address, profile photo, and behavior patterns. Additionally, the system can detect “bots,” often operated by organized crime gangs, which create profiles and engage real members in automated scripted conversations designed to elicit payment.

Already, through its quality assurance methods, the site has identified 20,000 scammers who get permanently blocked from Cupid’s communities each month.

Here are some expert tips on avoiding scams from Cupid.com’s Communications director, Sean Wood:

• Military pictures should turn on your scam radar warning. Some of the biggest scams occur from supposed soldiers in the armed services with a believable sob story and a desperate need to settle down.
• Try and keep the conversation going on the dating site messaging system rather than being dragged off to Skype or MSN, at least initially.
• Employ some basic conversational tests: If the person claims to be local, engage in a local-knowledge conversation (e.g. the weather, some nearby event, sports team, TV, or festival) and watch for mistakes. If the person is not claiming to be local, consider why he or she would want a long-distance relationship. Is there good reason, such as some particularly unusual common ground? If not, why aren’t they restricting themselves to people a bit nearer to their current location?
• Look out for words like “dear” and old-fashioned romantic language, as well as badly formed sentences.
•  Never, ever send money to anyone on a dating site, and especially avoid wiring cash.
• Always be on the lookout for get-rich-quick schemes. If it sounds too good to be true, then it’s probably a scam.

Word of caution: The FBI recently issued a warning about a different kind of online dating scam known as “ransomware.” It’s a virus that will make your computer inoperable until you hand over a payment. Read more about the ransom scam here. 

Have you ever been a victim of an online dating scam? To protect consumers, sites like romancescam.com are being set up for volunteers to post details of the counterfeit dating profiles they come across. Christine Davis, a design student who was duped by a fake profile, set up this particular site and wants to support others who have been targeted. Davis calls it the “Nigerian advance fee scam,” as it is thought that the majority of spammers are based in Western Africa.

Hacker image via Shutterstock

Sit down with execs from any tech company in America and ask them about their biggest obstacle to growth. 99 times out of a 100, they’ll tell you about their ongoing struggle to find technical talent.

Software engineers are more elusive than girls at a Silicon Valley tech conference. Similarly to finding a date, in a bid to attract the most sought-after engineers, recruiters typically turn to the Internet.

Entelo, a startup that is announcing its public beta today, has developed an algorithm that connects tech recruiters with likely candidates based on their online activity. It has indexed over 300 million social profiles — the “big data” technology can determine where tech savvy types like to go online and can pickup on their interests and expertise.

Most software engineers will exchange ideas and their most sophisticated code on sites like Github or Stackoverflow. They might also answer show off their knowledge on a Q&A site for Silicon Valley types like Quora. Meanwhile, designers can show off their skills on Dribbble (invitation-only show and tell for creatives) or online portfolio site, Behance.

When they are unfulfilled in a current job, chances are they’ll share content on these sites with a renewed vigor.

Founder Jon Bischke told me that the idea for the company originated during coffee meetings with fellow entrepreneurs. Most of them complained about the lack of engineering and sales candidates in their pipeline. The former entrepreneur in residence (EIR) at Battery Ventures said it was enough to convince him to start a company. Since then, Entelo has grown to 10 employees, and has raised an undisclosed sum in first round funding from Battery Ventures and Menlo Ventures.

Bischke is one of the few founders I have spoken with who has no problem recruiting engineers.

Entelo’s first customers have used the site to find talent in places you wouldn’t expect. For instance, if a college senior in Kansas posts some impressive code on Github, they might receive a call from one of Entelo’s customers the following week. Bischke told me that Github is a fan of Entelo as it encourages job-seekers to share their code in return for landing a dream job.

The algorithm prioritizes engineers at companies that are experiencing a rocky quarter, a low stock price, or lay-offs. This month, Zynga’s hot-shot mobile designer might appear at the top of the list. That very same designer would be pushed up further still in the rankings if they had recently posted some code or updated their status or location.

The company competes with “talent search engines” like TalentBin and Gild, as well as traditional online job boards. Its customers (primarily small and medium sized tech companies) pay $500 per month to use Entelo, which is a darn sight cheaper than the cost of hiring a recruiting agency. In future, Entelo may expand its offering to integrate with education startups like Coursera or Udacity, so tech companies can connect with the budding coders and designers who excelled at an online class.

---

Big Data and Predictive/Real-time Analytics startups: Are you looking to jumpstart development & accelerate market traction? Sign up for the SAP Startup Focus program to receive technology, support, resources and community to help you develop new applications on SAP HANA, a cutting edge database platform. Get started here, and enter promo code “VB2013″ on the form.

---

Sept. 9 - 10, 2013
San Francisco, CA

Early Bird Tickets on Sale

Imagine being able to predict when a war will break out. Well, scientists are trying, and they’re getting awfully close.

In 2010, when the WikiLeaks scandal broke, a group of friends gathered together at Bitly Inc in New York to try to bring order to the chaotic mess of data WikiLeaks had unleashed.

They used a simple code to extract dates and locations from about 77,000 unclassified reports from both simple stop-and-search operations and battles. This revealed unexpected hot spots: the Pakistani border, and the country’s main highway, which had experienced a series of violent outbreaks.

Recent advances in big data technologies have shown that we’re close to being able to predict exactly when a battle will break out. However there’s one thing that will stymie advances in this field: the inherent unpredictability of humans.

This year, when big data became a buzzword, the team reunited at Bitly’s offices to give the project another shot. This time, they teamed up with the brightest mathematical minds for a more audacious goal: a war zone prediction model. The researchers found a general pattern to the violence in Afghanistan, using it to determine whether an uprising would take place in each province, and its level of intensity.

The model worked with surprising accuracy and didn’t fail even when President Obama changed the rules of the game by sending in 30,000 additional troops.

The project, the results of which were published online by the Proceedings of the National Academy of Sciences in July, is just one small part of a growing movement to anticipate episodes of armed conflict using algorithmic computational techniques. Still, we have a long road ahead of us before this data is turned into actionable intelligence — a matter of life or death on the battlefield.

As Lt. Gen. Michael Oates, head of the Joint Improvised Explosive Device Organization, recently stated, “There is no shortage of data. There is a dearth of analysis.”

Iraq: the first ‘big data war’

Bitly’s research wasn’t the first time a group of renegade scientists brought the power of analytics to a war zone. For the better part of the decade, bringing big data to the battlefield has been the job of civilian researchers.

Cast your mind back to the spring of 2003, when four countries participated in the invasion of Iraq and succeeded in toppling the regime of Saddam Hussein in 21 days. At Oxford University, a young scholar had a theory, one that did not sit very well with centuries of political theory. He wondered whether wars share a single, predictable pattern.

Sean Gourley (pictured), a New Zealand-born graduate student, told me that “during one of those classic Oxford dinner conversations where you sit around these high tables, Harry Potter-style”  he butted heads with James Woolsey, the former director of the CIA.

“I had a hunch that there might be some strong, mathematical pattern that might emerge once we’d looked at Iraq,” he said. “No one had really done it before.”

Using a combination of reports from 130 news sources, SMS-based communications between freelance journalists and photographers stationed in Baghdad, plus any information from the frontlines he could get his hands on, he set to work on an algorithm.

“We were writing software to extract when people were dying and how they were dying …”

Gourley told me that he harbored strong reservations about the research. “My God … we were writing software to extract when people were dying and how they were dying,” Gourley said, admitting that he frequently considered throwing in the towel.

To make matters more complicated, reports from various news media often conflicted. As we know now, war reporting is notoriously inaccurate and is even less viable when comparing one armed conflict to another. In Afghanistan, for example, rural environs and a depleted number of reporters on the field led to less coverage than in the Iraq conflict. This remains a problem for researchers today.

Gourley told me they learned that the best approach was for humans and algorithms had to work hand-in-hand. By then, a team of physicists were working with him to see him through this crazy experiment, and they were investigating other conflict areas, including Sierra Leone. According to Gourley, the senior-level military personnel were keeping a watchful eye on their progress.

However, his frequent attempts to convince a contact at the Pentagon to hand over data fell on deaf ears. As a foreign national, he was not able to access official U.S. military reports.

This proved to be a blessing in disguise. At that time, the military’s analysts were typically trained in political science, not computer science, and their reports were spotty at best. “They couldn’t write a python script if you paid them,” Gourley joked. After the WikiLeaks scandal, he discovered that his datasets were superior to the U.S. military’s. “It turns out that we had 80 percent of what they had; they only had 70 percent of what we had,” he explained.

The results, published in Nature in 2009, found that insurgent wars follow an approximate power law, in which the frequency of attacks decreases with increasing attack size to the power of 2.5. That means that for any insurgent war, an attack with 10 casualties is 316 times more likely to occur than one with 100 casualties.

This may seem like a boring set of numbers, but for the first time, it revealed an underlying pattern to war. “It shows that there is something going on in the way these wars are fought that is common to all,” Neil F. Johnson, a physicist at the University of Miami who participated in the research, told Nature.

In the course of his research, he and his team collected data on 54,679 “violent events” reported in nine different conflicts, including those in Iraq, Afghanistan, Peru, and Colombia.

What’s next for the big data wars?

The Gourley and Bitly research are first steps in bringing objective quantitative analysis to realms that were once subjective. Big data will play a growing role in maintaining global security as the Department of Defense reshuffles budgets and priorities. According to Forbes, the amount of data from drones and other surveillance technology has risen 1,600 percent since 9/11.

To step up the research, the U.S. military recently made a $250 million bet on big data. In May, U.S. Secretary of Defense, Leon Panetta put forward a review of the country’s national defense, spotlighting information processing as a growing priority.

For Gourley, who ultimately left academia to form big data startup Quid, this research will have far-reaching consequences, and not just for the military. If these algorithms work, they may change the very nature of war.

Key Image via Shutterstock 

The startup AfterCollege is relaunching with a new matching algorithm to help seniors at 2,300 colleges and universities find the perfect entry-level job or internship. The company claims its technology was developed for graduating students to stand out in this fiercely competitive job market.

The new website is far more technical than the original, which was a simple online job board for college students that competed with sites like Indeed and CollegeBuilder. With its relaunch, the site is rebranding itself as a “LinkedIn for college students,” and users are encouraged to build out their professional profiles.

AfterCollege has a tough road ahead if it aims to take on LinkedIn and career networking startups like BranchOut that have spent years developing a presence at college campuses and marketing to students.

Most interesting is the built-in job search tools (a resume and cover-letter builder) and the patented-machine learning technology that matches seniors and recent grads with opportunities, based on their skill-sets, personal and academic interests.

The algorithm works by inputting a user’s academic affiliations (school, educational program, department, groups) and searching the database for relevant opportunities. AfterCollege has been around since 1999 and has been used by 3 million students, so it has access to a wealth of insider information.

For instance, the algorithm would know if the computer science department at San Francisco State has a different focus than the program at Stanford University. At Stanford, Human Computer Interaction (HCI) might be a strong interest for many students, while at SF State, students are skilled at Java or social applications — useful information for engineering recruiters and tech companies.

“The algorithm is also smart enough to distinguish between different programs that may be perceived to be similar,” explained AfterCollege CEO and founder Roberto Angulo. He started the company as a junior at Stanford. In 13 years, AfterCollege has connected students with some half a million jobs and internships.

AfterCollege also optimizes chances for student applicants as they can view jobs that graduates with similar skill-sets have succeeded in landing.

AfterCollege’s algorithm is a useful tool for employers, who fear they’ll miss out on the best candidates. No doubt, this will be a major money-making opportunity for AfterCollege; the site will likely begin selling targeted advertising to employers.

The startup raised its most recent round of funding in 2011, a sub-million-dollar sum from Flywheel Ventures.

Top Image via Shutterstock

Kaggle, a San Francisco-based startup that hosts data science competitions, has uncovered some disconcerting insights about human behavior in its two-year run. At times, its founders have been surprised by the accuracy of an algorithm, and the competitions continue to evoke controversy.

In short, data can be dangerous. I caught up with the company’s founder and CEO, Anthony Goldbloom, to find out more about recent data-driven discoveries that have rocked the boat.

1) “The Essay Scoring Competition”

Sponsor: Hewlett Foundation / Prize: $100,000
Goal: To get the computer to give an essay the same score a human grader would.

The idea was that by analyzing spelling and punctuation, as well as sentence structure, an algorithm could give an essay a reliable score, perhaps even more consistent than a human grader.

Martin O’Leary, a glacier scientist at the University of Michigan, was one of hundreds of competitors from around the world. He told Reuters that he discovered that human graders are rarely in agreement. They are swayed by irrelevant, aesthetic factors like how neatly a student writes. Unlike an algorithm, they award scores that seem random.

“The reality is, humans are not very good at doing this,” said Steve Graham, a Vanderbilt University professor who has researched essay grading techniques in an interview with Reuters.

A controversial study, indeed. Reflecting on the competition, Goldbloom was not initially convinced it could be done. ”I remember thinking: Are we going to be falling flat on our face? It’s really hard to take an essay and give it a grade,” he recalled. The biggest obstacle was to find a team with the requisite machine learning expertise and the ability to deal with unstructured data, including text.

As critics pointed out, it’s easy to outsmart an algorithm and optimize for success. In response to the New York Times’ article which advocated “obfuscating mellifluously” when facing a robograder (i.e. disarm with big words), Goldbloom said the algorithm would need to be refreshed at least once a year.

Key insight: An algorithm is no less reliable at scoring essays than the average teacher. 

2) “The Twitter psychopathy competition”

Sponsor: Online Privacy Foundation / Prize: $1,000
Goal: Can Twitter detect psychopathy? How about personality traits?

In two separate competitions, the goal was to analyze a person’s social media flow to detect personality type and risk for psychopathy. Can we get a sense for your personality in a single tweet?

The data, drawn from the tweets of 3,000 people, may surprise you. “It turns out that with Twitter data alone, we can go quite some way into figuring out someone’s personality,” said Goldbloom. The signals for psychopathy: good grammar, an angry tone, use of swear words and conjunctions.

Before their tweets were analyzed, the users completed a personality test. The algorithm immediately rooted out the most extroverted of the lot. Those who reference others in a reply are more likely to be extroverted. One competitor, a Japanese ad-targeting expert, found that openness is the personality trait that, ironically, is the most difficult to detect.

Are you giving more information about yourself than you realize? Goldbloom said this information could potentially be used by potential employers, as well as for targeted online advertising.

Key insight: With only 140 characters, data scientists and statisticians can get a strong sense for your personality. That’s fairly worrying, considering that this information could get into the wrong hands.

Featured Image via Shutterstock