A New Way to Predict Flight Delays

Coming home for thanksgiving break, I found myself stuck at the Richmond airport for 10 hours more than I had planned for. Our flight had been delayed, and I arrived in Boston at 2 in the morning instead of the original 4 PM. However, recently, researchers from Binghamton University in New York have created a revolutionary computer model that can accurately predict flight delays quickly. This model could potentially save travelers, like myself, hours of time before arriving at the airport.

The current method for predicting flight delays uses artificial neural network computer models, otherwise known as ANNs. Sina Khanmohammadi, leading the study, notes that their "proposed method is better suited to analyze datasets with categorical variables (qualitative variables such as weather or security risks instead of numerical ones) related to flight delays" (ScienceDaily). ANN models contain nodes that take in variables and predict outcomes through analyzing these variables. Through previous data, ANNs can look for patterns and are self- learning. Nevertheless, these networks can also be backfilled with data from previous flights and this can slow the process of flight delay predictions, causing delays to be reported later than desired.

This new research has created a new ANN to help airlines easily determine the relationships between input like weather or previous delays and output- the flight delays themselves. The method analyzes fourteen variables. When the team used the model to predict delays for hypothetical flights expected to land at JFK airport in New York City on January 21st, they found the model predicted delay length with "20 percent more accuracy" and took "40 percent less time" to reach those conclusions (EconomicTimes). They found that previous delays causing a chain reaction of delays were the biggest causes of late arrivals. In fact, that was one of the causes for my own delay last Tuesday. Not only will this new information be useful to fliers, but it can also help airline companies retain their customers and efficiently compensate for delayed flights.

References:
Content:
https://www.sciencedaily.com/releases/2016/11/161114103905.htm
http://economictimes.indiatimes.com/news/science/researchers-find-faster-way-to-predict-flight-delays/articleshow/55432529.cms
https://www.bupipedream.com/news/75742/auto-draft-171/

Images:
http://www.newsx.com/sites/default/files/styles/home_image/public/field/image/Researchers-find-faster-way-to-predict-flight-delays.jpg?itok=uqpIdhIG
http://image.slidesharecdn.com/final-141219125424-conversion-gate01/95/airline-flights-delay-prediction-2014-spring-data-mining-project-3-638.jpg?cb=1418993788

Insect Biobots and Drones Used to Map Disaster Areas

After a disaster, it can be dangerous for humans to enter an unknown and potentially unsafe area. This is why researchers at North Carolina State University developed the technology to use unmanned aerial vehicles, or drones, and insect biobots to enter these unknown places and inspect and map the areas. This technology would work well in areas such as a collapsed building, where gps can't be used. It can also help to locate stranded people and animals after disasters, to map out paths to reach them, and to determine how structurally safe certain paths may be compared to other ones.

The biobots these researchers have developed take the form of insects, like cockroaches. The job of the UAV's (the drones) is to contain and guide the biobots to the areas of interest. The drone hovers over the area, sending out a strong radio signal to contain the biobots. These bugbots are created by equipping cockroaches with sensors hooked up to their antennae and cerci, located on the front and back of the bug, respectively. Researchers can remotely stimulate these appendages to steer and drive the cockroaches. In this system, the biobots would be allowed to run freely through an area, with the beacon from the drone keeping the bugs in close proximity so they only map the desired area.

A specially designed algorithm takes the sensor data the biobots collect and translate it into a map of the area. The drone then changes locations, bringing along the bugbots, and the same process is repeated. The software can stitch multiple maps together, creating a comprehensive overview of the area. Co-author of the study Edgar Lobaton says that "We had previously developed proof-of-concept software that allowed us to map small areas with biobots, but this work allows us to map much larger areas and to stitch those maps together" (newatlas).

References:
Content:
http://newatlas.com/drones-cyborg-cockroaches-map-buildings/46510/
http://www.innovationtoronto.com/2016/11/using-uavs-and-insect-cyborgs-to-map-disaster-areas/
https://www.sciencedaily.com/releases/2016/11/161117104402.htm
Images:
https://images.sciencedaily.com/2016/11/161117104402_1_900x600.jpg
http://images.natureworldnews.com/data/images/full/36593/drones-and-insect-biobots-can-now-map-high-risk-disaster-areas.jpg

Cookies Are Not Just a Dessert!!

While cookies are delicious, you may have heard them in the context of computers before. Cookies are messages/ pieces of text that a web server passes to your web browser and stores on your hard disk when you visit an internet site. The web server can later go and retrieve the information, which can include content about a user's specific visit to a website (for example, items a customer has placed in a shopping cart), or content a user has volunteered about themselves or entered previously into a form (like a name, address, or credit card number). The messages stored on your machine are in small text files that include name-value pairs. The website that created the cookie is the only one that can access and read it later, and it cannot access any other information on your hard disk.

You can access and view all of the cookies stored on your computer by looking at the directory where a web browser holds this information. You can also manipulate the settings on your computer to have control over the process of websites storing name-value pairs on your computer.

The website How Stuff Works provides a good overview of how the cookie data moves:
First, when you visit any website, your web browser will contact a website's server for the specific page you are requesting. Your browser will simultaneously look on your own hard disk for a text file that the website has previously set. If it finds a cookie file, it will send this information to the server alongside the request for the page. If you have never visited a website before, you most likely will not have any cookie data stored on your computer for that site. In this case, the server will not receive any cookie data and will usually create a new ID for you, the new user, in it's database, and sends this information to your machine which stores it on your computer. The web server can alter the data it stores on your hard disk whenever you visit it's website.

Websites use cookies to remember what state your browser was in, gather information of how many people are visiting their site and what they are doing on it, to store user preferences or customize a site for a specific visitor, and more. There has been some concerns about issues with privacy and security that come with the introduction of cookies. The main issues are that some companies can sell your information to other third party companies that want to sell you similar things, making customer targeting precise. Also, some infrastructure providers, like DoubleClick, create cookies visible on multiple sites, gathering information about you from many different places and developing a detailed profile of you. These companies serve ads on so many sites, and it's scary to some people that they have the power to link these detailed profiles of the users to names and personalize the profiles.

References:
Content:
http://computer.howstuffworks.com/cookie1.htm
https://kb.iu.edu/d/agwm
https://en.wikipedia.org/wiki/HTTP_cookie

Images:
https://cldup.com/_d6DG7yyR8.png
http://www.webopedia.com/DidYouKnow/Images/051807_cookies001.jpg
https://www.grc.com/cookies/browser4.png

Secret Messages in Dance Music??

Have you ever wondered if there was a hidden message in the song you were dancing to? Well, there very well might be. A Polish man named Krysztof Szczypiorski from the Warsaw University of Technology has created a new form of cryptography. He has created an algorithm that can encode and decode secret messages in dance music. These messages are virtually undetectable by the human ear. Szczypiorski has named his program "StegIbiza." "Steg" stands for steganography, which is basically hiding hidden messages in plain sight. Audio steganography is not a new concept; throughout history notes have been used to spell out hidden messages in different ways. However, Szczypiorski's form of steganography is unique.

Szczypiorski uses variation in tempo embedded in various club dance style songs to encrypt his messages. The tempo changes are so subtle that professional musicians, when asked to listen for the message, could only detect the changes when the tempo was altered by two and three percents. The tempo changes act as a "digital morse code;" when the tempo is sped up, it represents a dot, while slowing down the tempo represents a dash. Szczypiorski's algorithm can detect the changes in tempo, decoding the message.

While this algorithm may not be revolutionary or change any lives, it is a pretty cool way to send hidden messages through music.

References:
Content:
http://blogs.discovermagazine.com/d-brief/2016/08/24/algorithm-weaves-secret-messages-into-dance-music/#.WBzhUOErKRs
http://idmmag.com/news/hidden-messages-dance-music/
http://www.pcworld.com/article/3109281/a-new-algorithm-can-hide-messages-in-your-favorite-dance-music.html

Images:
http://www.pcworld.com/article/3109281/a-new-algorithm-can-hide-messages-in-your-favorite-dance-music.html
http://blogs.discovermagazine.com/d-brief/files/2016/08/shutterstock_268477412.jpg

Computer Program that Replicates Handwriting!

UCL computer scientists have recently developed a computer program that can analyze a person's handwriting and accurately reproduce it. They named the program "My Text in Your Handwriting." This software could be useful because it allows you to write a note in your handwriting without even touching a pen and paper. Furthermore, it could be useful for stroke victims who may not be able to write legibly anymore, or for the replication of the handwriting of someone who is deceased.

The program works by scanning a paragraph or more of text in a user's handwriting and then converting it into a thin line. The computer works to break down this writing, looking for different letters and their positions. It also looks for other elements of writing like "splits" (where the line changes from a letter into ligatures, the extra parts that comprise joined handwriting) and "links" (two separate marks that are part of the same letter, like dotting an i). The computer constantly references previously scanned examples, but applies a degree of randomness in its replication of the writing so that the same letter shape combinations are not used every time. The computer also figures out the appropriate spacing to place between each letter and the size and position of the letters on the page. The computer mimics the pen line texture and color you were using by adding texturing.

The results of this software are pretty believable. The UCL scientists asked members of a group to decide which envelopes out of a pile of handwritten and software generated addresses were produced by a computer, and the group chose incorrectly 40% of the time.

While fonts and the ability to convert writing into text or text into writing have been developed previously, the ability to write in personalized handwriting is unprecedented. The only problem with this software is that it could be used for forgery for signatures or documents.

References:

Content:

http://phys.org/news/2016-08-programme-replicates.html

http://visual.cs.ucl.ac.uk/pubs/handwriting/

https://www.engadget.com/2016/08/12/a-computer-program-that-can-replicate-your-handwriting/

Images:

http://visual.cs.ucl.ac.uk/pubs/handwriting/img/results.jpg

What is the Cloud, Exactly??

I know that I've heard the term "could computing" many times, but I've never fully understood what it means. This post explains what the cloud is exactly, some of its benefits and uses,

In general, the cloud is a network of servers, each with a different function. Servers can be used to run applications, deliver a "service," or store data. The cloud is a metaphor for the internet; cloud computing means storing and accessing data and programs through the internet instead of a local storage area like your computer's hard drive. Local storage is fast and easy, but for those on the local network. With the cloud, cloud computing can be done anywhere, anytime with an online connection. For example, pictures you take on your phone are stored on the phone's internal hard drive; however, when you upload a picture to Instagram or Facebook, you are storing the information on the cloud. Chances are, you encounter and use the cloud all the time. Applications and services like Google Drive, iCloud, Dropbox, even Netflix all use the cloud. Some services work in the cloud and store a synced version of your files online but also sync those files with local storage. When your data is stored "in the cloud," that basically means that it is stored on a remote server owned by a company like Apple or Google. Services that use the cloud pay these companies a membership fee to provide this remote storage area.

Benefits:
Companies used to have to buy all of their own storage and hardware equipment. With the introduction of the cloud, companies only have to pay for what they use. This makes it easy for a company to scale up or down, and they don't have to worry about staying up to date with storage equipment. This creates a company that is more cost effective and efficient. Cloud computing also makes it easy for a company to back up its data; it provides a way out if a company loses all of its local data or its local servers crash. With cloud computing, automatic software updates are also an added benefit, as well as the ability to work from anywhere, anytime. Documents are easily shareable among multiple people, and this means that the environmental impact of paper use can be easily reduced.

Arguments against the cloud:
The intellectual property line for information stored in the cloud can be very blurred: it can be hard to tell what information you own, and what information is owned by the company storing it. A provider could have a strong claim on the data you create in the cloud vs. the data you upload to the cloud. Furthermore, your internet service provider/ telecom company controls your access, and companies can continue to make you pay for doing more and more things because they have control over your data. Also, if your information access is unavailable or offline, you will, in most cases, be unable to work with data stored in the cloud. Although most cloud service providers implement great security standards and certifications, using external service providers always opens the door to security risks. You provide the provider with access to important data, and because that data is stored on the server of someone else, you are giving your trust (and money) to them to protect it.

References:

Content:
http://mashable.com/2013/08/26/what-is-the-cloud/#SZ5hx7iBZkqZ
http://gizmodo.com/what-is-the-cloud-and-where-is-it-1682276210
http://www.pcmag.com/article2/0,2817,2372163,00.asp

Images:
http://www.pcmag.com/article2/0,2817,2372163,00.asp
https://www.cloudbuckit.com/images/blog/56/cloud-computing.png

Mind Controlled Robotic Arm that Allows Paralyzed Man to Feel Again!!!

Imagine a world in which paralyzed men and woman could regain their sense of touch. Recently, researchers at the University of Pittsburgh and the University of Pittsburgh Medical Center collaborated to demonstrate a technology that allows a paralyzed patient to experience the sensation of touch through a robotic arm that he controls with his brain. The patient's name is Nathan Copeland, and he injured his spinal cord in a car accident in 2004, leaving him paralyzed from the chest down and without the sensation of touch since his accident.

For the experiment, four micro-electrode arrays were implanted in the sensory cortex of Copeland's brain last spring. This is where the neurons that control hand movement and touch are located. Before the surgery, imaging techniques were used to identify the exact location of Copeland's brain that corresponded to each of his fingers and his palm. These micro-electrode arrays allowed Copeland to be connected to the BCI or Brain Computer Interface.

BCIs work through using electrodes to connect the brain to a computer program. The electrodes in the brain measure the small differences in the voltage between the neurons, the nerve cells in our brains that transmit electric signals. This signal is then amplified and filtered. After that, the signal is read as input into a computer program and interpreted. This can also work the other way around; the computer can convert a signal, such as one from a robotic arm, into the voltage necessary to trigger neurons. The signals are transmitted to the electrodes in the right area of the brain and, when received, the neurons fire in that area.

In Copeland's case, he is the first human to have experienced a sensation of touch through a robotic arm. When the doctor touched each of the fingers on the robot's right hand, Copeland described the experience as a natural sensation of touch instead of tingling, and was able to identify correctly the exact location of each of the touches on his hand.

References:

Content:
http://www.nbcnews.com/health/health-news/brain-chip-helps-paralyzed-man-feel-his-fingers-n665881
https://www.washingtonpost.com/news/to-your-health/wp/2016/10/13/in-a-medical-first-brain-implant-allows-paralyzed-man-to-feel-again/
http://www.upmc.com/media/NewsReleases/2016/Pages/bci_scitransl-lms.aspx
http://computer.howstuffworks.com/brain-computer-interface.htm

Images:
https://www.sciencedaily.com/releases/2016/10/161013151356.htm
http://media3.s-nbcnews.com/j/newscms/2016_41/1749331/161013-robotic-hand-mn-1104_bd769511961914a5a51eb1d3316f3b2f.nbcnews-ux-2880-1000.jpg

The Real Deal with Credit Card Chips

Have you ever gone to check out at a store and swiped your credit card, just to be told that you need to use the chip reader? And then you hold up the line for ten minutes because you can't figure out how long to leave it in for or what to press? What is the point of this new chip technology, and how does it work?

Chip technology has been around for about 30 years, although the United States has recently broadly adopted the technology in the last five years. In 2013, hackers stole credit and debit card information from 40 million Target customers, and there were also massive data breaches elsewhere in the retail industry. These hackers, who were never caught, were believed to have installed malware onto the checkout card readers that stole data each time a customer swiped his or her card. This was the incentive for companies to start adopting the more secure microchips instead of the magnetic strips on credit cards that are easy to hack.

Microchips are commonly referred to as EMV smart chips, which stands for Europay, MasterCard and Visa: the companies that created this technology. The difference between microchips and magnetic strips is that instead of embedding credit card numbers and user information into the strip, it is contained in the little chip in the card. In addition to storing data, the computer chip can process data. When a user inserts the chip into the card reader, the chip and the reader essentially have an encrypted conversation to authenticate the card. The chip uses cryptography to protect the data. When using a magnetic strip, the swipe of the card simply outputs the card number and expiration date to the reader, making these cards easy to counterfeit.

In the US, the most common type of chip cards currently are the "chip and signature" cards, which require a signature to authorize the charge. In contrast, many European countries use "chip and PIN" cards, which require a four-digit PIN to authorize the charge. The "chip and PIN" cards are more secure because a signature can be easily forged.

References:

Content:
https://en.wikipedia.org/wiki/EMV
http://money.howstuffworks.com/personal-finance/debt-management/chip-and-pin-credit-cards3.htm
http://blog.credit.com/2014/10/how-does-chip-pin-actually-work-97454/


Images:
https://static01.nyt.com/images/2016/08/04/business/04cardchip4/04cardchip4-master768.jpg
https://assets.nerdwallet.com/blog/wp-content/uploads/2014/01/best-emv-chip-credit-cards.jpg

Speech Recognition Software

Speech recognition Software is incredibly prevalent in society today, with uses that include for computer dictation software, automated call centers, digital voices for those with disabilities, and smartphone "voices" like Siri. What these all have in common is that they work by somehow analyzing the "phones," or sound packets that our voices generate, in different ways. Phones are the real bits of sound that we speak. Speech recognition software is highly complex because of the difficulty of its task. There are many obstacles it runs into: how to separate slurred words, how to differentiate between voices and pitch, how to put homophobes in context, a sentence that is misheard, etc.

There are a options for how different speech recognition softwares work. The first option is pattern matching: where a spoken word is recognized entirely. The second option is Pattern and feature analysis: where a work is broken into parts and each part is analyzed and recognized for key features like the vowels. Another option is language modeling and statistical analysis: where speech recognition speed and accuracy are improved by a knowledge of grammar and the probability of certain sounds in a specific order. The last option is artificial neural networks: brain-like computer models that are trained to recognize patterns.

In general, speech recognition softwares first operates by using an analog-to-digital converter (ADC) to translate the vibration waves that you generate by speaking into digitized sound by measuring the wave at intervals. It removes unwanted noise and background noise and separates the pitch, or frequency bands, of the speech. It adjusts the sound volume to a steady level, and divides the signal into pieces (the phones) into known phonemes, or tiny sounds, in a specific language. There are about 44 phonemes in the english language. It runs the phoneme through a statistical model to determine what the user was saying and usually outputs it as text.


References:

Content:
https://en.wikipedia.org/wiki/Speech_recognition
http://electronics.howstuffworks.com/gadgets/high-tech-gadgets/speech-recognition.htm
http://www.explainthatstuff.com/voicerecognition.html
https://msdn.microsoft.com/en-us/library/hh378337(v=office.14).aspx
http://scienceline.org/2014/08/ever-wondered-how-does-speech-to-text-software-work/

Images:
http://efv-solutions.com/wp-content/uploads/2015/09/Home.png
http://rethink-wireless.com/wp-content/uploads/2016/05/Siri_3359038b.jpg

Shopping Cart Software

While it is easy to mindlessly scroll through a store's website online and add things to a virtual shopping cart, online shopping is a pretty complex process. Online shopping allows customers to browse through a store's catalog, view products and add selections to their carts without being locked in to purchasing the items. If a customer decides to purchase an item later, he or she can "check out" that specific item from the cart, enter payment information, and the product will be shipped to the destination entered.

If an online store plans to display and sell over 50 products, it will usually use a store-building system as opposed to an online ordering system. The store-building system and the online ordering system both handle the product pages for each of the products, but the store-building system handles orders on the same site instead of redirecting a customer to another web page to pay and complete the transaction.

The shopping cart feature allows the customer to indicate the quantity and specifications of the products he or she wants and save them to his or her personal cart. The store-building system will usually show the running total of items in the cart. In order for the shopping cart program to differentiate between shoppers, a variety of systems are used. Most sites use cookies: small computer files that identify a shopper and remain on a shopper's hard disk while on the website.  Another method includes software that can read the temporary IP number of a customer that is assigned by the customer's internet service provider upon logging onto the internet.

When it is time to check out, the software will usually calculate a customer's total cost, including shipping fees and taxes. Because taxes differ by states, some carts allow tax calculation by zip code. Shipping calculations are based on a number of specifications, including sales total, weight, and number of items in the order. Some sites have a flat rate for shipping any order. To get real-time authorization of credit cards, a secure payment gateway is needed. The payment gateway software is usually provided by an outside service, not the online store itself. Finally, in order for the transaction to be secure, a Secure Sockets Layer or SSL secure connection is needed between the customer and the website to encrypt communications and prevent hackers from stealing personal information. When the merchant retrieves and downloads a placed order, there also needs to be a secure SSL connection or another secure method to make sure credit card information is not exposed.

References:

Content:
https://en.wikipedia.org/wiki/Shopping_cart_software
http://www.practicalecommerce.com/articles/100112-cart_intro
https://www.impactbnd.com/blog/online-shopping-cart-work

Images:
https://secure.fatcow.com/images/user-guide/screenshot-ecom-checkout.jpg
https://compass.ups.com/uploadedImages/Artical_Area/Articles/1112_ShoppingCart_Article.jpg

Point of Service Systems

The point of service is the time and place where a retail transaction is completed. Point of sale or point of service systems (POS) are useful services in the retail industry. They allow a merchant to easily calculate and display the amount owed by a customer, make exchanges, process orders, and provide receipts for transactions without any need for a price tag. POS systems' functionality differs for different business depending on a business models and processes. Modern POS systems are reliable, easy to use, low cost, and run many different functions with a high and consistent operating speed. They can aid in the growth of a business by acting as a strategic service center.

Most POS systems involve highly complex programming to create because of the extensiveness of the operations involved. A POS machine may need to process a products price, balance, average cost, quantity sold, description, and department. While one of the main components of a POS system involves taking inventory, it could also be used to add stock, create barcodes, hold membership records, track a customer's purchase history, and run reports. These systems have to account for sales and the constantly changing prices of products. Overall, each POS system holds a wide range of capabilities, but requires a certain amount of customization based on what an end-user needs.

A POS system requires a high level of accuracy in its code because of the high demands of the performance of the system. It is constantly receiving submissions and retrievals of data to process transactions, and most transactions are monetary. An inventory list is maintained on the system's hard drive. In order for a POS system to be a successful product, much testing and debugging is required before the product goes on the market so the system can be as efficient as possible.

In recent times, cloud computing has allowed POS systems to be accessed directly from an internet browser instead of installed locally. This allows user data to be stored remotely instead of on a local server so data can be accessed from anywhere, and the system can be continually remotely updated. However, POS systems that use the cloud cannot function without the vendor, forcing a company to be less self-sufficient and secure.

References:

Content:
https://en.wikipedia.org/wiki/Point_of_sale
https://www.entrepreneur.com/article/77960
http://www.mobiletransaction.org/how-pos-system-work/
http://smallbusiness.chron.com/pos-systems-work-41627.html

Images:
http://www.fourlane.com/product_images/uploaded_images/qbpos-sys-land.jpg
http://mms.businesswire.com/media/20130620005353/en/373576/5/Heckler-with-SK-screen-1.jpg

The Fitbit: Wearable Technology

In a society that is increasingly concerned about healthy lifestyles and fitness, the Fitbit is a widely popular form of wearable technology. The Fitbit company claims to encourage users to meet their daily exercise goals and to make fitness a regular part of their lifestyles by allowing users to track their progress. A Fitbit is a small device that can be worn around the wrist or attached to clothing that tracks a user's activity, exercise, food, weight, and sleep. It syncs the statistics wirelessly and automatically from a person's Fitbit device to his or her phone or computer. This allows a user to track his or her progress online through charts and graphs.

The web-based profile allows a user to create an account with his or her height, age, gender, weight, and it also allows a user to log meals online. Users can set short term goals (like steps per day) or long term goals (like weight loss) and track their progress. They can also connect with social media to motivate other users as well.

Fitbit uses a three-dimensional accelerometer system to gather data that tracks motion and the intensity of that motion. Next, the Fitbit software uses algorithms to convert the accelerometer data into usable information. The company has refined these algorithms by experimenting and comparing Fitbit's accuracy with other fitness trackers. The accelerometer tracks energy expenditure and motion, and the algorithms convert this data into information on number of calories burned, steps taken and distance moved. To track sleep, the Fitbit gathers data on how much a user moves at night and then uses algorithms to convert this data to time and quality of sleep. Every time a user passes within fifteen feet of his or her wireless base station, the Fitbit will sync the data from the device onto the online profile.

There have been many studies testing how accurate the Fitbit is at measuring activity. Overall, Fitbit is fairly accurate: more accurate at measuring steps and distance, but a little less reliable for counting calories.


References:

Content:
http://electronics.howstuffworks.com/gadgets/fitness/fitbit2.htm
https://www.fitbit.com/fitscience
https://techcrunch.com/2011/10/03/the-fitbit-ultra-more-fitbit-goodness-same-fitbit-package/
http://berkeleysciencereview.com/fit-fitbit/

Images:
http://az616578.vo.msecnd.net/files/2016/05/28/636000086421781248-470868957_fitbit-banner-mobile.jpg
https://static0.fitbit.com/simple.b-cssdisabled-png.hd53c63eab7cc1316845e4b4084fbc504.pack?items=%2Fcontent%2Fassets%2Fcharge%2Fimages%2Fapp-dashboard%2Fscreen%2Fweb_Next_Gen_Dash%402x.png

The Computer Science Side of Pandora Internet Radio

The Internet Radio service named Pandora claims it can guide you, the listener, to new (or old) music that you enjoy based on a song or artist that you like. Upon navigating to the website (www.pandora.com) the user types in a song, artist, or genre of music to the web-based music player. This creates a "radio station" that will in turn play music Pandora deems similar to the song or artist originally entered.

Pandora is based on something called the Music Genome Project. Each song in the Genome database has been analyzed by a trained music analyst using up to 450 musical characteristics to determine the genetic makeup of the song. Some of these traits include a song's arrangement, melody, rhythm, beat, tempo, and voice. When the user enters a song to create a "radio station," Pandora runs an algorithm to compare every song in the Music Genome database to the one entered, identifying songs with similar characteristics. These songs are the ones played next on the "radio station." The user can further refine the station by pressing the "thumbs up" or "thumbs down" buttons on a particular song, providing input that tweaks the algorithm so that the station plays more or fewer songs similar to the one playing, respectively.

Pandora stores the listener's user data in his or her computer's Flash local storage. This helps Pandora recognize when the same listener returns to the site. Pandora also stores songs that have already been played in a specific listening session in the users computer's Flash local storage so the same songs do not get chosen multiple times in a session.

Aside from using user input to refine his or her personal radio station, Pandora uses the "thumbs up" and "thumbs down" feedback from millions of listeners to further refine its algorithm. This allows Pandora to make better informed projections about what songs might fit a listener's taste.

References:

Content:
http://computer.howstuffworks.com/internet/basics/pandora3.htm
https://www.fastcompany.com/1808123/how-computers-curators-and-users-create-pandoras-playlists
http://www.nytimes.com/2009/10/18/magazine/18Pandora-t.html
http://www.nytimes.com/2014/01/05/technology/pandora-mines-users-data-to-better-target-ads.html

Images:
http://images.dailytech.com/nimage/22015_large_Pandora.png
https://images.iskysoft.com/mac-audio-recorder/pandora-site-signing-in-problems-810-524.jpg