The Less Scary Applications of Artificial Intelligence: Computer Vision
August 3, 2016
The article on The Christian Science Monitor titled Shutterstock’s Reverse Image Search Promises a Gentler Side of AI provides a glimpse into computer vision, or the way a computer assesses and categorizes any image into its parts. Shutterstock finds that using machine learning to find other images similar to the first is a vast improvement, because rather than analyzing keywords, AI analyzes the image directly based on exact colors and shapes. The article states,
“That keyword data, while useful for indexing images into categories on our site, wasn’t nearly as effective for surfacing the best and most relevant content,” says Kevin Lester, vice president of engineering at the company, in a blog post. “So our computer vision team worked to apply machine learning techniques to reimagine and rebuild that process.”
The neural network has now examined 70 million images and 4 million video clips in its collection.”
In addition, the company plans to expand the search feature to videos as well as images. Jon Oringer, CEO and founder of Shutterstock, has a vision of endless possibilities for this technology. The article points out that this is one of the clearly positive effects of AI, which gets a bad rap, perhaps not unfairly, given the potential for autonomous weapons and commercial abuse. So by all means, let’s use AI to recognize a cat, like Google, or to analyze images.
Chelsea Kerwin, August 3, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Internet Sovereignty, Apathy, and the Cloud
December 21, 2015
The OS News post titled Dark Clouds Over the Internet presents an argument that boils down to a choice between international accord and data sharing agreement, or the risk of the Internet being broken up into national networks. Some very worked up commenters engaged in an interesting discussion that spanned government overreaching, democracy, data security, privacy, and for some reason, climate change. One person summarized their opinion thusly:
“Best policy: don’t store data with someone else. There is no cloud. It’s just someone else’s computer.”
In response, a user named Alfman replied that companies are to blame for the current lack of data security, or more precisely, people are generally to blame for allowing this state of affairs to exist,
The privacy issues we’re now seeing are a direct consequence of corporate business models pushing our data into their central silos. None of this is surprising except perhaps how willing users have been to forgo their own privacy. Collectively, it seems that we are very willing to give up our rights for very little in exchange… makes it difficult to achieve critical mass around technologies promoting data independence.”
It is hard to argue with the apathy factor, with data breaches occurring regularly and so little being done by individuals to protect themselves. Good thing these commenters have figured it all out. Next up, solving climate change.
Chelsea Kerwin, December 21, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Where’s the Finish Line Enterprise Search?
September 16, 2015
What never ceases to amaze me is that people are always perplexed when goals for technology change. It always comes with a big hullabaloo and rather than complaining about the changes, time would be better spent learning ways to adapt and learn from the changes. Enterprise search is one of those technology topics that sees slow growth, but when changes occur they are huge. Digital Workplace Group tracks the newest changes in enterprise search, explains why they happened, and how to adapt: “7 Ways The Goal Posts On Enterprise Search Have Moved.”
After spending an inordinate amount of explaining how the author arrived at the seven ways enterprise search has changed, we are finally treated to the bulk of the article. Among the seven reasons are obvious insights that have been discussed in prior articles on Beyond Search, but there are new ideas to ruminate about. Among the obvious are that users want direct answers, they expect search to do more than find information, and understanding a user’s intent. While the obvious insights are already implemented in search engines, enterprise search lags behind.
Enterprise search should work on a more personalized level due it being part of a closed network and how people rely on it to fulfill an immediate need. A social filter could be applied to display a user’s personal data in search results and also users rely on the search filter as a quick shortcut feature. Enterprise search is way behind in taking advantage of search analytics and how users consume and manipulate data.
“To summarize everything above: Search isn’t about search; it’s about finding, connecting, answers, behaviors and productivity. Some of the above changes are already here within enterprises. Some are still just being tested in the consumer space. But all of them point to a new phase in the life of the Internet, intranets, computer technology and the experience of modern, digital work.”
As always there is a lot of room for enterprise search improvement, but these changes need to made for an updated and better work experience.
Whitney Grace, September 16, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Computers Learn Discrimination from Their Programmers
September 14, 2015
One of the greatest lessons one take learn from the Broadway classic South Pacific is that children aren’t born racist, rather they learn about racism from their parents and other adults. Computers are supposed to be infallible, objective machines, but according to Gizmodo’s article, “Computer Programs Can Be As Biased As Humans.” In this case, computers are “children” and they observe discriminatory behavior from their programmers.
As an example, the article explains how companies use job application software to sift through prospective employees’ resumes. Algorithms are used to search for keywords related to experience and skills with the goal of being unbiased related to sex and ethnicity. The algorithms could also be used to sift out resumes that contain certain phrases and other information.
“Recently, there’s been discussion of whether these selection algorithms might be learning how to be biased. Many of the programs used to screen job applications are what computer scientists call machine-learning algorithms, which are good at detecting and learning patterns of behavior. Amazon uses machine-learning algorithms to learn your shopping habits and recommend products; Netflix uses them, too.”
The machine learning algorithms are mimicking the same discrimination habits of humans. To catch these computer generated biases, other machine learning algorithms are being implemented to keep the other algorithms in check. Another option to avoid the biases is to reload the data in a different manner so the algorithms do not fall into the old habits. From a practical stand point it makes sense: if something does not work the first few times, change the way it is done.
Whitney Grace, September 14, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Web Sites Going The Way Of The Dodo
July 24, 2015
Apps are supposed to replace Web sites, but there is a holdup for universal adoption. Search Engine Watch explains why Web sites are still hanging tight and how a new Google acquisition might be a game changer: “The Final Hurdle Is Cleared-Apps Will Replace Web Sites.” The article explains that people are “co-users” of both apps and classic Web sites, but online browsers are still popular. Why is that?
Browsers are universal and can access any content with a Web address. Most Web sites also do not have an app counterpart, so the only way to access content is to use the old-fashioned browser. Another issue is that apps cannot be crawled by search engines, so they are left out of search results. The biggest pitfall for apps is that they have to be downloaded in order to be accessed, which takes up screen space and disk space.
A startup has created a solution to making apps work faster:
“Agawi has developed a technology to stream apps, just like Netflix streams videos. Instead of packaging the entire app into a single, large file for the user to download, the app is broken up into many small files, letting users interact with small portions of the app while the rest of it is downloading. In the short term, it appears that Google wants to deploy Agawi for users try an app before downloading the full version.”
Google acquired Agawi, but do not expect it to be accessible soon. Google enjoys putting its own seal of approval on all acquisitions and making sure it works well. Mobile device usage is increasing and more users are moving towards using them over traditional computers. Search marketers will need to be more aware than ever about how search engines work with apps and encourage clients to make an app.
Whitney Grace, July 24, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
-
- Subscribe to Beyond Search
Feature archive
News archive
-
