Word Embedding Captures Semantic Relationships
November 10, 2016
The article on O’Reilly titled Capturing Semantic Meanings Using Deep Learning explores word embedding in natural language processing. NLP systems typically encode word strings, but word embedding offers a more complex approach that emphasizes relationships and similarities between words by treating them as vectors. The article posits,
For example, let’s take the words woman, man, queen, and king. We can get their vector representations and use basic algebraic operations to find semantic similarities. Measuring similarity between vectors is possible using measures such as cosine similarity. So, when we subtract the vector of the word man from the vector of the word woman, then its cosine distance would be close to the distance between the word queen minus the word king (see Figure 1).
The article investigates the various neural network models that prevent the expense of working with large data. Word2Vec, CBOW, and continuous skip-gram are touted as models and the article goes into great technical detail about the entire process. The final result is that the vectors understand the semantic relationship between the words in the example. Why does this approach to NLP matter? A few applications include predicting future business applications, sentiment analysis, and semantic image searches.
Chelsea Kerwin, November 10, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
No Search Just Browse Images on FindA.Photo
March 2, 2016
The search engine FindA.Photo proves itself to be a useful resource for browsing images based on any number of markers. The site offers a general search by terms, or the option of browsing images by color, collection (for example, “wild animals,” or “reflections”) or source. The developer of the site, David Barker, described his goals for the services on Product Hunt,
“I wanted to make a search for all of the CC0 image sites that are available. I know there are already a few search sites out there, but I specifically wanted to create one that was: simple and fast (and I’m working on making it faster), powerful (you can add options to your search for things like predominant colors and image size with just text), and something that could have contributions from anyone (via GitHub pull requests).”
My first click on a swatch of royal blue delivered 651 images of oceans, skies, panoramas of oceans and skies, jellyfish ballooning underwater, seagulls soaring etc. That may be my own fault for choosing such a clichéd color, but you get the idea. I had better (more various) results through the collections search, which includes “action,” “long-exposure,” “technology,” “light rays,” and “landmarks,” the last of which I immediately clicked for a collage of photos of the Eiffel Tower, Louvre, Big Ben, and the Great Wall of China.
Chelsea Kerwin, March 2, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Big Data Is so Last Year, Data Analysts Inform Us
February 1, 2016
The article on Fortune titled Has Big Data Gone Mainstream? asks whether big data is now an expected part of data analysis. The “merger” as Deloitte advisor Tom Davenport puts it, makes big data an indistinguishable aspect of data crunching. Only a few years ago, it was a scary buzzword that executives scrambled to understand and few experts specialized in. The article shows what has changed lately,
“Now, however, universities offer specialized master’s degrees for advanced data analytics and companies are creating their own in-house programs to train talent in data science. The Deloitte report cites networking giant Cisco CSCO -4.22% as an example of a company that created an internal data science training program that over 200 employees have gone through. Because of media reports, consulting services, and analysts talking up “big data,” people now generally understand what big data means…”
Davenport sums up the trend nicely with the statement that people are tired of reading about big data and ready to “do it.” So what will replace big data as the current mysterious buzzword that irks laypeople and the C-suite simultaneously? The article suggests “cognitive computing” or computer systems using artificial intelligence for speech recognition, object identification, and machine learning. Buzz, buzz!
Chelsea Kerwin, February 1, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Bing Uses Image Search for Recipes
December 8, 2015
Recipe websites have become the modern alternative to traditional cookbooks, but finding the perfect recipe through an Internet search engine can be tedious. LifeHacker informs us that Bing is now using image search technology to help users whittle down the results in, “Find Recipes by Image in Bing’s Image Search.” Writer Melanie Pinola describes how it works:
“When you look up ‘baked ziti’ or ‘roast turkey’ or any other food-related term and then go to Bing’s images tab, photos that you can access recipes for will have a chef’s hat icon, along with a count of how many sites use that image. Click on the image to see the recipe(s) related to the image and load them in your browser. You’ll save some time versus click through to every recipe in a long list of search results, especially if you’re thinking of making something that looks a particular way, such as bacon egg cups.”
So remember to use Bing next time you’re hunting for a recipe online. Image search tech continues to improve, and there are many potential worthwhile uses. We wonder what it will be applied to next.
Cynthia Murrell, December 8, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Predictive Search Tries to Work with Videos
October 14, 2015
Predictive search is a common feature in search engines such as Google. It is more well-known as auto-complete, where based on spelling and keyword content the search engine predicts what a user is searching for. Predictive search speeds up the act of searching, but ever since YouTube became the second biggest search engine after Google one has to wonder if “Does Video Enhance Predictive Search?” asks Search Engine Watch.
Search engine and publisher of travel deals Travelzoo created a video series called “#zootips” that was designed to answer travel questions people might search for on Google. The idea behind the video series was that the videos would act as a type of predictive feature anticipating a traveler’s needs.
“‘There’s always push and pull with information,’ says Justin Soffer, vice president of marketing at Travelzoo. ‘A lot of what search is, is people pulling – meaning they’re looking for something specific. What videos are doing is more of a push, telling people what to look for and showing them things.’ ”
Along with Travelzoo, representatives from SEO-PR and Imagination Publishing also agree that video will enhance video search. Travelzoo says that video makes Web content more personal, because an actual person is delivering it. SEO-PR recommends researching keywords with Google Trends and creating videos centered on those words. Imagination Publishing believes that video content will increase a Web site’s Google ranking as it ranks media rich pages higher and there is an increase in voice search and demand for how-to videos.
It is predicted that YouTube’s demand as a search engine will increase more content will be created for video. If you understand how video and predictive analytics work, you will have an edge in future Google rankings.
Whitney Grace, October 14, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Why Are Ads Hiding Themselves
June 25, 2015
The main point of an advertisement is to get your attention and persuade you to buy a good or service. So why would ads be hiding themselves in a public venue? Gizmodo reports that in Russia certain ads are hiding from law enforcement in the article: “This Ad For Banned Food In Russia Itself From The Cops.” Russian authorities have banned imported food from the United States and European Union. Don Giulio Salumeria is a Russian food store that makes its income by selling imported Italian food, but in light of the recent ban the store has had to come up with some creative advertising:
“Websites are already able to serve up ads customized for whoever happens to be viewing a page. Now an ad agency in Russia is taking that idea one step further with an outdoor billboard that’s able to automatically hide when it spots the police coming.”
Using a camera equipped with facial recognition software programmed to recognized symbols and logos on officers’ uniforms, the billboard switches ads from Don Giulio Salumeria to another ad advertising a doll store. While the ad does change when it “sees “ the police coming, they still have enough time to see it. The article argues that the billboard’s idea is more interesting than anything. It then points out how advertising will become more personally targeted in the future, such as a billboard recognizing a sports logo and advertising an event related to your favorite team or being able to recognize your car on a weekly commute, then recommending a vacation. While Web sites are already able to do this by tracking cookies on your browser, it is another thing to being tracked in the real world by targeted ads.
Whitney Grace, June 25, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
MIT Discover Object Recognition
June 23, 2015
MIT did not discover object recognition, but researchers did teach a deep-learning system designed to recognize and classify scenes can also be used to recognize individual objects. Kurzweil describes the exciting development in the article, “MIT Deep-Learning System Autonomously Learns To Identify Objects.” The MIT researchers realized that deep-learning could be used for object identification, when they were training a machine to identify scenes. They complied a library of seven million entries categorized by scenes, when they learned that object recognition and scene-recognition had the possibility of working in tandem.
“ ‘Deep learning works very well, but it’s very hard to understand why it works — what is the internal representation that the network is building,’ says Antonio Torralba, an associate professor of computer science and engineering at MIT and a senior author on the new paper.”
When the deep-learning network was processing scenes, it was fifty percent accurate compared to a human’s eighty percent accuracy. While the network was busy identifying scenes, at the same time it was learning how to recognize objects as well. The researchers are still trying to work out the kinks in the deep-learning process and have decided to start over. They are retraining their networks on the same data sets, but taking a new approach to see how scene and object recognition tie in together or if they go in different directions.
Deep-leaning networks have major ramifications, including the improvement for many industries. However, will deep-learning be applied to basic search? Image search still does not work well when you search by an actual image.
Whitney Grace, June 23, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
-
- Subscribe to Beyond Search
Feature archive
News archive
-
