Typically the effectiveness of recommender systems is determined by statistical accuracy metrics of the algorithm used such as MAE (Mean Absolute Error). However, Kirsten Swearingen and Rashmi Sinha argue interaction design is equally important in determining recommender system effectiveness. Performing an analysis of music recommender systems, researchers have discovered that there are two separate models for Recommender System success. The first is in terms of ecommerce (user's indication they will buy music) and the other is usefulness (how users are helped to explore musical tastes). Eleven systems were tested (including the likes of Amazon, MovieCritic, Media Unbound and CDNow). System recommendations provided by 6 of those systems are compared to recommendations provided by their friends. Study 1 involved 20 participants and Study 2 involved 12, all of which are regular Internet users in the 19 to 44 age range. Users provided input to the system and received a set of recommendations. Users were then asked to rate 10 recommendations from each system, evaluating aspects such as liking, action towards item (buy/download/do nothing), transparency (do they understand why system recommended that item) and familiarity (any previous experience of the item). users were also asked to rate the systemas a whole on a number of dimensions: usefulness, trustworthiness, and ease of use. At the end of the session, users were asked to name the system they preferred and explain their rationale.
Findings
The goal of most recommender systems is to replace or at least augment the social recommendation process. Study results showed that users preferred recommendations made by their friends versus online systems but their is a high level of overall satisfaction finding them useful in suggesting items that users had not previously heard of. Users like the breadth that online systems offer, allowing them the unique opportunity to exploer their tastes and learn about new items. Effective recommenders inspire trust and users are willing to provide more input to the system in return for more accurate recommendations. Designers often try to balance ease of use while enhancing accuracy. Of the participants studied, 67% didn't think the 4-20 input ratings amazon requires is sufficient to generate accurate recommendations. Conversely, MediaUnbound requiers 34 input ratings and 75% thinks that number is just right. Users also commented on the rating process noting some mechanisms such as genre selection, labeling your favorite artist, and rating scales are either too restrictive or redundant and boring. Users did like the rating bar/slider scale since they could click anywhere to indicate their degree of liking. They expressed interest in varying the rating process to one that is engaging and offers mixed questions and continuous feedback. Participants also like receiving information about recommended items. Identifying why it was recommended, when it was released, album covers and reviews by others is useful. Adjustments were made to RatingZone's Quick Picks to offer my detail about recommendations and they noticed a 20% increase in usefulness! In addition, users like and prefer to buy previously familiar recommendations. It helps build trust. For example, participants said 72% of Amazon's, 60% of MediaUnbounds and 45% of MoodLogic's recommendations were familiar. There is a greater willingness to buy familiar than unfamiliar recommended items. this makes sense since a familiar item is a less risky purchase decision. Users expressed a willingness to buy only 7% of the items recommended by mediaUnbound. While users show a preference for familiar items they do express frustration over recommendations that were albums by the same artists that the users had input into the system. Amazon might remind users about a favorite song not heard recently but it did not help users expand their tastes in new directions. MediaUnbound includes a silder bar for users to indicate how familiar the music suggested should be and users participants stated they liked this feature.
Again, while the algorithm used to generate the recommendation is useful in determining effectiveness, interaction factors must be equally weighed. Depending on the definition of success used, ecommerce techniques may be more important than usefulness and vice versa. Of course if the goal is to receive the best of both worlds, a hybrid system that uses the strengths of each approach is ideal!
Interaction Design for Recommender Systems (2002)
HUBRIS: Human Benchmarking of Recommender Systems (2002)
Kirsten Swearingen
Rashmi Sinha
Marti Hearst
Thursday, April 10, 2008
Thursday, April 3, 2008
Songkick Concert Recommender
So while album/track purchases may be down drastically, it seems artist may be recouping through live concerts. In 2007, concert ticket sales earned about $9 billion worldwide, thats pretty impressive!! Songkick is attempting to capitalize on the "big boom" by offering a centralized location to search for concerts information (date, locations and ticket prices). Without registering, visitors can input up to three artist/bands at a time and the service will provide concert dates, if available, or they will recommend other artist/bands which will be touring in a city near you and which they assume you are likely to enjoy!
One of the features songkick totes is its ability to look at your computer's music library (iTunes, Winamp and Windows Media Player) and build recommendations. Songkicker is a downloadable plug-in which automatically scans your entire music library and adds these artists into your Tour Tracker -- without you ever having to do anything! The Songkicker automatically runs in the background whenever you play music. Every track you play is Songkicked to us and appears in your "Recently Songkicked" section. Though it seems to free up a lot of time and lessens the amount of manual user input, I don't feel very comfortable with "tracking devices" on my PC....thats just my personal opinion though!
So How Does it Work?
Good question!! The most Ive been able to decipher is this....Any mention of music on the web is a data point for recommendations. Their recommendation engine doesnt generate suggestions from the user base like Last.fm or through careful analysis like my favorite toy, Pandora, but instead it crawls websites like Wikipedia and music blogs to pick up related artisted based on positive or negative associations between the bands. Through a combination of including anything about music on the internet coupled with using "expert" critical opinion from blogs and music publications, the technology is able to infer similarities between artists, compare this to your personal music taste and recommend concerts that you might actually enjoy. I wish I could get a better grasp of what depicts positive and negative or better details about how "similarities" between artists are identified but I'm unable to locate the specifics. It just launched on March 18, 2008 so maybe more details will be released in the future!
Supposedly the real payoff is buying tickets, allowing you to find the cheapest tickets for these shows. Songkick provides direct links to ticket inventory from 16 vendors across the U.S. and U.K.
As a closing point - for artist and band bloggers, Songkick, offers these individuals a way to make extra money with a widget that identifies bands on tour mentioned on their blogs, and inserts tour information and ticket vendor links that can be tracked for referrals.
Did I Test It Out?
Of course I did! The first three artists I input were individuals who I know are touring! I input Jay-Z and Mary J Blige (they will be in Greensboro NC on Saturday) and then Kanye West (he's coming to Charlotte on May 2). Jay-Z and Mary didnt return any results, not even the concert in Greensboro but Kanye did appear. I switched the artists and added 3 more artists: Lauryn Hill, The Roots and Pharrell Williams. Both The Roots and Pharrell are coming to Charlotte but I didn't get any results for them either. So I decided to signup (username/password: prsrecommend) and I played around a little more. I selected John Mayer, Maroon 5 and Avril Lavigne. I didn't know any of these artists were touring and actually found their dates in Charlotte so that was pretty neat. I am still extremely interested in knowing how similar artists are derived because they did seem to do a decent job of suggesting artists that I know and enjoy. Perhaps it's based on popularity and if they're scanning the web, I would bet money that's how it's determined.....maybe I'll continue investigating!!
One of the features songkick totes is its ability to look at your computer's music library (iTunes, Winamp and Windows Media Player) and build recommendations. Songkicker is a downloadable plug-in which automatically scans your entire music library and adds these artists into your Tour Tracker -- without you ever having to do anything! The Songkicker automatically runs in the background whenever you play music. Every track you play is Songkicked to us and appears in your "Recently Songkicked" section. Though it seems to free up a lot of time and lessens the amount of manual user input, I don't feel very comfortable with "tracking devices" on my PC....thats just my personal opinion though!
So How Does it Work?
Good question!! The most Ive been able to decipher is this....Any mention of music on the web is a data point for recommendations. Their recommendation engine doesnt generate suggestions from the user base like Last.fm or through careful analysis like my favorite toy, Pandora, but instead it crawls websites like Wikipedia and music blogs to pick up related artisted based on positive or negative associations between the bands. Through a combination of including anything about music on the internet coupled with using "expert" critical opinion from blogs and music publications, the technology is able to infer similarities between artists, compare this to your personal music taste and recommend concerts that you might actually enjoy. I wish I could get a better grasp of what depicts positive and negative or better details about how "similarities" between artists are identified but I'm unable to locate the specifics. It just launched on March 18, 2008 so maybe more details will be released in the future!
Supposedly the real payoff is buying tickets, allowing you to find the cheapest tickets for these shows. Songkick provides direct links to ticket inventory from 16 vendors across the U.S. and U.K.
As a closing point - for artist and band bloggers, Songkick, offers these individuals a way to make extra money with a widget that identifies bands on tour mentioned on their blogs, and inserts tour information and ticket vendor links that can be tracked for referrals.
Did I Test It Out?
Of course I did! The first three artists I input were individuals who I know are touring! I input Jay-Z and Mary J Blige (they will be in Greensboro NC on Saturday) and then Kanye West (he's coming to Charlotte on May 2). Jay-Z and Mary didnt return any results, not even the concert in Greensboro but Kanye did appear. I switched the artists and added 3 more artists: Lauryn Hill, The Roots and Pharrell Williams. Both The Roots and Pharrell are coming to Charlotte but I didn't get any results for them either. So I decided to signup (username/password: prsrecommend) and I played around a little more. I selected John Mayer, Maroon 5 and Avril Lavigne. I didn't know any of these artists were touring and actually found their dates in Charlotte so that was pretty neat. I am still extremely interested in knowing how similar artists are derived because they did seem to do a decent job of suggesting artists that I know and enjoy. Perhaps it's based on popularity and if they're scanning the web, I would bet money that's how it's determined.....maybe I'll continue investigating!!
Songkick: Live Music Lovers Will Love This
Songkick's Concert Recommendation Engine: It Goes to 11
Thursday, March 27, 2008
Is like-i-like really what-i-like?
like-i-like.org is a movie recommendation system that oddly enough recommends movies! lol It learns your movie taste, from ratings provided by the user, and makes personalized predictions specially for you. It offers a feature which allows you to find "movie soulmates", individuals who have similar ratings and preferences. You can even see how many votes your "soulmate" has provided and see how they voted on a particular movie!! These soulmates can also be contacted to ask for their opinion on movies that you have yet to view. According to the program's FAQ on achieving the best (most precise) predictions, users should be objective in their rankings. Predictions appear after you rate at least 10 items, but going over 20 ranked items results in much moer accurate forecast.
So i registered to use the system and after rating 30+ movies, i question whether their recommendations are in fact "what-i-like"? It appears the movie database in still in a building stage because a number of the movies that I searched for to vote on where not found. After rating 6 movies, predictions began to appear on how I would rate the movie. Of the following 6 movies that I rated, 3 of the predictions were accurate (when rounded) and 3 where within 1 or 2 points. So after giving 12 votes, I decided to see what type of recommendations the system would offer. Viewing only the top 10, all of the movies suggested were ranked 9.9 and I have to admit I've never seen or even heard of them :-( Perhaps, this is why they suggest ranking at least 20 movies. So onward I went. I rated another 9 movies and 3 of the predictions were accurate (when rounded) and again the rest were within a point or two. What struck me as odd during this period is I performed a search for Spiderman...to my amazement, the movie didn't appear!! Right or wrong, I question the validity of any system that doesn't have at least the first Spiderman in its database....call me crazy but that's one of the best movies, all time!! So by the end of this phase I've rated a total of 21 songs and wanted to take a look at the recommendations again. The system predicted I would rate each movie a 10 but with the exception of Dirty Harry-which i vaguely remember b/c of the title though I dont remember watching it, none look even remotely familiar. I've never claimed to be a movie connoisseur so I like what I like and on rare occassion, I step outside of that realm and land on something interesting that may have never caught my eye. Even with that said, I guess I was expecting to see more movies that I've seen or heard about and would enjoy watching. I mean I've rated some of my all-time favorites at this point and they should give a pretty good indication of what I like but either I'm the odd ball and other users don't agree with my taste or the system isn't as knowledgeable as I would like. Wanting to go just another step further, I rated 10 more movies. On 2 of the 10, the system didn't offer any predictions at all. 5 of the remaining 8 were accurate and the last 3 were within a few points of accuracy. Worth noting, the system predicated i would rank Malibu's Most Wanted, which I thought was a very poor movie though I admittedly laughed maybe twice, at least a 5.5 and when rounded is a 6 which is slightly good when i actually gave it a 3. Conversely, Training Day which I think is a phenomenal movie, masterpiece even, was only rated 8.9...I wish I could see the brains of the operation to give it a piece of my mind too! lol So I decided to check out the recommendations again, hoping for the best, and still nothing!! Two of the 10 movies recommended, previously appeared in the list and that lets me know either the system thinks it's right and that I would REALLY REALLY like this movie or the system has not be able to infer anything from my additional ratings :-( I didn't recognize a single title in the entire list....perhaps I just expect too much!! lol
Remi A - signing off!
(BTW - because of the ranking/rating system used, like-i-like uses collaborative filtering to make recommendations)
like-i-like.org
username: prs-tester
password: prs-tester
So i registered to use the system and after rating 30+ movies, i question whether their recommendations are in fact "what-i-like"? It appears the movie database in still in a building stage because a number of the movies that I searched for to vote on where not found. After rating 6 movies, predictions began to appear on how I would rate the movie. Of the following 6 movies that I rated, 3 of the predictions were accurate (when rounded) and 3 where within 1 or 2 points. So after giving 12 votes, I decided to see what type of recommendations the system would offer. Viewing only the top 10, all of the movies suggested were ranked 9.9 and I have to admit I've never seen or even heard of them :-( Perhaps, this is why they suggest ranking at least 20 movies. So onward I went. I rated another 9 movies and 3 of the predictions were accurate (when rounded) and again the rest were within a point or two. What struck me as odd during this period is I performed a search for Spiderman...to my amazement, the movie didn't appear!! Right or wrong, I question the validity of any system that doesn't have at least the first Spiderman in its database....call me crazy but that's one of the best movies, all time!! So by the end of this phase I've rated a total of 21 songs and wanted to take a look at the recommendations again. The system predicted I would rate each movie a 10 but with the exception of Dirty Harry-which i vaguely remember b/c of the title though I dont remember watching it, none look even remotely familiar. I've never claimed to be a movie connoisseur so I like what I like and on rare occassion, I step outside of that realm and land on something interesting that may have never caught my eye. Even with that said, I guess I was expecting to see more movies that I've seen or heard about and would enjoy watching. I mean I've rated some of my all-time favorites at this point and they should give a pretty good indication of what I like but either I'm the odd ball and other users don't agree with my taste or the system isn't as knowledgeable as I would like. Wanting to go just another step further, I rated 10 more movies. On 2 of the 10, the system didn't offer any predictions at all. 5 of the remaining 8 were accurate and the last 3 were within a few points of accuracy. Worth noting, the system predicated i would rank Malibu's Most Wanted, which I thought was a very poor movie though I admittedly laughed maybe twice, at least a 5.5 and when rounded is a 6 which is slightly good when i actually gave it a 3. Conversely, Training Day which I think is a phenomenal movie, masterpiece even, was only rated 8.9...I wish I could see the brains of the operation to give it a piece of my mind too! lol So I decided to check out the recommendations again, hoping for the best, and still nothing!! Two of the 10 movies recommended, previously appeared in the list and that lets me know either the system thinks it's right and that I would REALLY REALLY like this movie or the system has not be able to infer anything from my additional ratings :-( I didn't recognize a single title in the entire list....perhaps I just expect too much!! lol
Remi A - signing off!
(BTW - because of the ranking/rating system used, like-i-like uses collaborative filtering to make recommendations)
like-i-like.org
username: prs-tester
password: prs-tester
Thursday, March 20, 2008
Males Getting an Electronic Personal Stylist?
The Galeria Kaufhof in Essen, Germany has found a new, intriguing way to use RFID tags. Typically, the RFID craze was surrounded around tracking items from suppliers to checkout. But, with new efforts from Checkpoint Systems, Impinj and Reva Systems, men buying clothes in RFID supported stores will get automatic suggestions!! Cool huh?! According to Baseline, when men go into the dressing room to try on a suit, a "smart mirorr" tells them what kind of shirt or tie they should buy with it. The notion is that business intelligence systems like smart mirrors and smart shelves can be used to increase sales while providing viable options to customers real-time.
Concept?
These smart mirrors have RFID readers which can determine what apparel has been brought into the change room or is wearing or holding at the table, based on their attached RFID tag. The mirror then displays complementary clothing choices or accessories. Just in conjunction with "smart shelves", the user is presented with only those options (ie sizes, styles and colors) that are available in-store.
How it works?
Thats a great question. Unfortunately, most of the documentation that I found online is written in German, revealing one of my many weaknesses in that Im not bilingual! So I'm left to infer possibilities. Being that the tags are placed on each item in the store, I gather there's some type of item or content based filtering behind the scenes. I would also suggests there may perhaps be some collaborative filtering as well given suggestions can be derived from past customers buying patterns and what items they pair together. I also wonder if perhaps some suggestions are entered into the system before involving any additional filtering. From the company standpoint, I'm sure they have designers and clothes buyers that are familiar with "what goes well together" so perhaps they use that upfront knowledge to offer recommendations and continue to develop those as more data is gathered. I definitely believe the suggestions rests on some type of hybrid recommendation system and am eager to determine if my assumptions are accurate, perhaps I can find a German speaking assistant between now and the end of the semester :-)
So far this technology is only available to males as the belief is they need more help. I would only offer that since women tend to shop more, financial benefits may be seen more readily by implementing this technology for females. Just my thoughts!
RFID tags help you to choose clothes (October 16, 2007)
Retailer Uses RFID to Help the Sartorially Challenged Sex (October 4, 2007)
World's First End-to-End UHF Item-Level RFID Shopping Experience for METRO Group (September 20, 2007)
METRO Group - The Spirit of Commerce
Concept?
These smart mirrors have RFID readers which can determine what apparel has been brought into the change room or is wearing or holding at the table, based on their attached RFID tag. The mirror then displays complementary clothing choices or accessories. Just in conjunction with "smart shelves", the user is presented with only those options (ie sizes, styles and colors) that are available in-store.
How it works?
Thats a great question. Unfortunately, most of the documentation that I found online is written in German, revealing one of my many weaknesses in that Im not bilingual! So I'm left to infer possibilities. Being that the tags are placed on each item in the store, I gather there's some type of item or content based filtering behind the scenes. I would also suggests there may perhaps be some collaborative filtering as well given suggestions can be derived from past customers buying patterns and what items they pair together. I also wonder if perhaps some suggestions are entered into the system before involving any additional filtering. From the company standpoint, I'm sure they have designers and clothes buyers that are familiar with "what goes well together" so perhaps they use that upfront knowledge to offer recommendations and continue to develop those as more data is gathered. I definitely believe the suggestions rests on some type of hybrid recommendation system and am eager to determine if my assumptions are accurate, perhaps I can find a German speaking assistant between now and the end of the semester :-)
So far this technology is only available to males as the belief is they need more help. I would only offer that since women tend to shop more, financial benefits may be seen more readily by implementing this technology for females. Just my thoughts!
RFID tags help you to choose clothes (October 16, 2007)
Retailer Uses RFID to Help the Sartorially Challenged Sex (October 4, 2007)
World's First End-to-End UHF Item-Level RFID Shopping Experience for METRO Group (September 20, 2007)
METRO Group - The Spirit of Commerce
Thursday, February 21, 2008
Tailrank (They claim to track the hottest news in the blogosphere -- maybe not so much)
Tailrank is a service that finds the best content from millions of blogs and provides a custom ranking specific to the user. Tailrank also offers a mobile version and users can import blog (OPML) subscriptions to build a personalized reading list.
Tailrank, founded by Kevin Burton, filters news according to your interests and those of others. Tailrank allows each user to share news and blogs with other members of the site. Tailrank finds the internet's hottest channels by indexing over 24M weblogs and feeds per hour. Tailrank uses Spinn3r, which is a blog spider that can be specialized using your own behavior insteading of creating a separate crawler. The Spinn3r API is free to researchers. The hottest stories are discovered by tracking conversations between blogs. Tailrank takes into consideration linking behavior, the text of the post, links in common with other users, text relevance, weblog ranking, past performance, and various other factors for recommendations. The ranking algorithm finds weblogs that are highly linked and discussed links and citations. The biggest challenge facing Tailrank is the amount of data that needs to be processed and the need to keep that data consistent within a distributed system, about 52TB of raw blog content a month and continuous processing of 160Mbits of IO.
Though Tailrank states that it primarily uses collaborative filtering, as noted on it's website, I'm led to believe there is some content filtering involved as well because it searches the text of the post and uses that information to make recommendations.
What I've found even more interesting is a blog that Tailrank Goes Blank, But Nobody Notices! It stated in July that the technology section of the site had no news whatsoever, it was completely blank and at one point the rest of the site went blank too. Michael Arrington, the blogger, emailed Kevin Burton who responded they were in the middle of an infrastructure upgrade. Michael noted that shortly after the site went live again but the stories were days old. Interestingly Michael write, "the fact that the technology ssection was down for weeks and no one seemed to notice or write about it suggests that the site isn't being read regularly by very many people...Tailrank is basically a showcase for the technology behind his other startup, Spinn3r, which provides blog indexing and ranking services to other sites. If Tailrank can't stay up to date with the news, how can partners rely on the underlying technology?" Which is indeed a valid point!
Tailrank has 5 tabs: General, Technology, Politics, Entertainment and Video. When I registered on the site, I attempted to click on all the tabs and actually the Entertainment section was not available. After checking back 20 minutes later it was available. I also noticed that while Tailrank claims to spider only "blogs", an article I clicked on regarding the total lunar eclipse directed me to CNN.com/technology. So I imagine all websites are being scanned for news at this point. Just to test out this theory, I clicked on a few more links regarding the lunar eclipse and was directed to articles on msnbc, usatoday and thestar.com. I also noticed the timeframes on the suggested readings. The most recent article was 16 hours ago and the oldest was 24 hours ago, with only 8 recommendations on the page. While the articles were relevant, I was rather disappointed to see "old" items. Given the lunar eclipse, ,which I watched and found completely fascinating, happened last night, I expected to see articles that were written less than 12 hours ago at best since the eclipse ended a little after midnight. The other thing I just noticed is that I clicked on the entertainment tab and was offered information regarding the lunar eclipse though most of the sites that the articles were snipped from where found under the technology section...and there is a technology tab on Tailrank so I wonder why the articles weren't placed there instead. Another interesting point that I noticed is the somewhat biasness of the recommendations. For instance, I clicked on the Politics tab and received an overwhelming amount of information regarding John McCain. Apparently, the majority of users are avid McCain followers (which explains the recommendations if we're using a collaborative system) or Tailrank has incorporated some type of baseline information into the system that automatically thinks/suggests certain items such as McCain. I'm actually not a McCain follower so maybe thats why Im a little disenchanted with the recommendations but I can't imagine I'm the only Tailrank user that has a different opinion be it with Politics, Technology or Entertainment...ok, I think that's enough nitpicking for now! lol
Tailrank.com (user info: oneup/oneup)
Tailrank Architecture - Learn How to Track Memes Across the Entire Blogosphere (Nov 19, 2007)
Tailrank Goes Blank, but Nobody Notices (Jul 2, 2007)
Tailrank: A Social News Recommendation and Filtering System Gets a New Look (Jan 12, 2006)
Tailrank - Social News Recommendation Engine (Nov 10, 2005)
Tailrank - A Tool for the Long Tail (Sept 21, 2005)
Tailrank - About Us
Tailrank, founded by Kevin Burton, filters news according to your interests and those of others. Tailrank allows each user to share news and blogs with other members of the site. Tailrank finds the internet's hottest channels by indexing over 24M weblogs and feeds per hour. Tailrank uses Spinn3r, which is a blog spider that can be specialized using your own behavior insteading of creating a separate crawler. The Spinn3r API is free to researchers. The hottest stories are discovered by tracking conversations between blogs. Tailrank takes into consideration linking behavior, the text of the post, links in common with other users, text relevance, weblog ranking, past performance, and various other factors for recommendations. The ranking algorithm finds weblogs that are highly linked and discussed links and citations. The biggest challenge facing Tailrank is the amount of data that needs to be processed and the need to keep that data consistent within a distributed system, about 52TB of raw blog content a month and continuous processing of 160Mbits of IO.
Though Tailrank states that it primarily uses collaborative filtering, as noted on it's website, I'm led to believe there is some content filtering involved as well because it searches the text of the post and uses that information to make recommendations.
What I've found even more interesting is a blog that Tailrank Goes Blank, But Nobody Notices! It stated in July that the technology section of the site had no news whatsoever, it was completely blank and at one point the rest of the site went blank too. Michael Arrington, the blogger, emailed Kevin Burton who responded they were in the middle of an infrastructure upgrade. Michael noted that shortly after the site went live again but the stories were days old. Interestingly Michael write, "the fact that the technology ssection was down for weeks and no one seemed to notice or write about it suggests that the site isn't being read regularly by very many people...Tailrank is basically a showcase for the technology behind his other startup, Spinn3r, which provides blog indexing and ranking services to other sites. If Tailrank can't stay up to date with the news, how can partners rely on the underlying technology?" Which is indeed a valid point!
Tailrank has 5 tabs: General, Technology, Politics, Entertainment and Video. When I registered on the site, I attempted to click on all the tabs and actually the Entertainment section was not available. After checking back 20 minutes later it was available. I also noticed that while Tailrank claims to spider only "blogs", an article I clicked on regarding the total lunar eclipse directed me to CNN.com/technology. So I imagine all websites are being scanned for news at this point. Just to test out this theory, I clicked on a few more links regarding the lunar eclipse and was directed to articles on msnbc, usatoday and thestar.com. I also noticed the timeframes on the suggested readings. The most recent article was 16 hours ago and the oldest was 24 hours ago, with only 8 recommendations on the page. While the articles were relevant, I was rather disappointed to see "old" items. Given the lunar eclipse, ,which I watched and found completely fascinating, happened last night, I expected to see articles that were written less than 12 hours ago at best since the eclipse ended a little after midnight. The other thing I just noticed is that I clicked on the entertainment tab and was offered information regarding the lunar eclipse though most of the sites that the articles were snipped from where found under the technology section...and there is a technology tab on Tailrank so I wonder why the articles weren't placed there instead. Another interesting point that I noticed is the somewhat biasness of the recommendations. For instance, I clicked on the Politics tab and received an overwhelming amount of information regarding John McCain. Apparently, the majority of users are avid McCain followers (which explains the recommendations if we're using a collaborative system) or Tailrank has incorporated some type of baseline information into the system that automatically thinks/suggests certain items such as McCain. I'm actually not a McCain follower so maybe thats why Im a little disenchanted with the recommendations but I can't imagine I'm the only Tailrank user that has a different opinion be it with Politics, Technology or Entertainment...ok, I think that's enough nitpicking for now! lol
Tailrank.com (user info: oneup/oneup)
Tailrank Architecture - Learn How to Track Memes Across the Entire Blogosphere (Nov 19, 2007)
Tailrank Goes Blank, but Nobody Notices (Jul 2, 2007)
Tailrank: A Social News Recommendation and Filtering System Gets a New Look (Jan 12, 2006)
Tailrank - Social News Recommendation Engine (Nov 10, 2005)
Tailrank - A Tool for the Long Tail (Sept 21, 2005)
Tailrank - About Us
Thursday, February 14, 2008
Smart Phone Tells You What To Do
Articles Written Between Sept and Oct 2007
Dai Nippon Printing (DNP) and Palo Alto Research Center Inc (PARC) have been collaborating since 2005 to develop a context and activity aware system that recommends information about "local area" activities, such as shopping and dining, movies and bookstores or concerts and bars, matching the consumer's location, time of day and personal tastes. PARC named the software Magitti (derived from two early design concepts: a magic scope and a digital graffiti system). When this software is installed on your GPS-enabled phone Magitti starts to suggest what to do in your area. When an individual accesses their phone they will instantly see a list of recommendations. If it's noon, the osftware might suggest local restaurants. If it's 3PM, it might recommend a nearby boutique for shopping. If it's 9PM, a list of pubs might appear. Of particular interest to mobile recommendation, is its usefulness in unfamiliar areas to the user. For tourists, vacationers, business travel or even exploring unchartered areas of your own city, mobile recommendations prevent users from having to wander foreign streets and ask someone passing by for directions or their opinion, which has the potential to be a nightmare!
Interestingly, "Magitti pulls GPS data from a user's phone, as well as text messages, emails and information about events saved in the phone's calendar, and uploads it to a server, along with the user's search terms", says Kurt Partridge (a researcher at PARC). He states text messages are important bits of information because they often include data about future plans. If for instance a person gets a text message suggesting sushi, the software will put recommendations for sushi and Japanese restaurants higher on the list. The mere mention of text messages, emails and calendar notes being monitored and stored have raised several privacy concerns. PARC states these messages are only kept for a short amount of time but ultimately, there's a trade-off between privacy and convenience. Bo Begole, a co-leader on the project, remarked that the analysis happens on the handset and not on the servers at the company.
The software uses artificial-intelligence algorithms to make tailored recommendations. After reading several articles, I would infer this recommender system uses a hybrid model (context, demographic and collaborative). By making recommendations from text messages, emails and calendar notes and comparing them with location, time of day, and other personal tastes it seems to rely on context and demographic. I also read that collaborative filtering is used to recommend things that others with similar tastes like and allows people to input their own ratings and reviews.
Testers have noted the software works more often than not. One tester detailed at 11:30 am (pacific time) the system offered nearby lunch restaurants, a home furnishings store and a gym, noting it was rather easy to expand or limit the distance of suggestions and the type of cuisine. They likened the application to the Apple iPhone but commented the interface isn't nearly as slick. While the software is functional, there are still technical problems that have not been solved such as category ambiguity. Shopping could mean farmer's market or Macy's. Eating could mean sitting down at a restaurant or picking up a sandwich at the grocery store or enjoying a meal at home so the semantics of keywords still need to be analyzed for each individual in an effort to provide more accurate results.
Magitti will go through public trials with young adults in Tokyo, Spring 2008. Deployment is scheduled next year in Japan but it is unlikely this software will be sold in Europe or in the U.S.
DNP, PARC Jointly Develop Recommender System for Mobile Terminals
A Phone That Tells You What To Do
From PARC, The Mobile Phone As Tour Guide
Smart Phone Suggests Things To Do
Snapshots of Application
Dai Nippon Printing (DNP) and Palo Alto Research Center Inc (PARC) have been collaborating since 2005 to develop a context and activity aware system that recommends information about "local area" activities, such as shopping and dining, movies and bookstores or concerts and bars, matching the consumer's location, time of day and personal tastes. PARC named the software Magitti (derived from two early design concepts: a magic scope and a digital graffiti system). When this software is installed on your GPS-enabled phone Magitti starts to suggest what to do in your area. When an individual accesses their phone they will instantly see a list of recommendations. If it's noon, the osftware might suggest local restaurants. If it's 3PM, it might recommend a nearby boutique for shopping. If it's 9PM, a list of pubs might appear. Of particular interest to mobile recommendation, is its usefulness in unfamiliar areas to the user. For tourists, vacationers, business travel or even exploring unchartered areas of your own city, mobile recommendations prevent users from having to wander foreign streets and ask someone passing by for directions or their opinion, which has the potential to be a nightmare!
Interestingly, "Magitti pulls GPS data from a user's phone, as well as text messages, emails and information about events saved in the phone's calendar, and uploads it to a server, along with the user's search terms", says Kurt Partridge (a researcher at PARC). He states text messages are important bits of information because they often include data about future plans. If for instance a person gets a text message suggesting sushi, the software will put recommendations for sushi and Japanese restaurants higher on the list. The mere mention of text messages, emails and calendar notes being monitored and stored have raised several privacy concerns. PARC states these messages are only kept for a short amount of time but ultimately, there's a trade-off between privacy and convenience. Bo Begole, a co-leader on the project, remarked that the analysis happens on the handset and not on the servers at the company.
The software uses artificial-intelligence algorithms to make tailored recommendations. After reading several articles, I would infer this recommender system uses a hybrid model (context, demographic and collaborative). By making recommendations from text messages, emails and calendar notes and comparing them with location, time of day, and other personal tastes it seems to rely on context and demographic. I also read that collaborative filtering is used to recommend things that others with similar tastes like and allows people to input their own ratings and reviews.
Testers have noted the software works more often than not. One tester detailed at 11:30 am (pacific time) the system offered nearby lunch restaurants, a home furnishings store and a gym, noting it was rather easy to expand or limit the distance of suggestions and the type of cuisine. They likened the application to the Apple iPhone but commented the interface isn't nearly as slick. While the software is functional, there are still technical problems that have not been solved such as category ambiguity. Shopping could mean farmer's market or Macy's. Eating could mean sitting down at a restaurant or picking up a sandwich at the grocery store or enjoying a meal at home so the semantics of keywords still need to be analyzed for each individual in an effort to provide more accurate results.
Magitti will go through public trials with young adults in Tokyo, Spring 2008. Deployment is scheduled next year in Japan but it is unlikely this software will be sold in Europe or in the U.S.
DNP, PARC Jointly Develop Recommender System for Mobile Terminals
A Phone That Tells You What To Do
From PARC, The Mobile Phone As Tour Guide
Smart Phone Suggests Things To Do
Snapshots of Application
Thursday, February 7, 2008
Personal Recommendation Software Predicts Consumer Choice
Jeffrey O'Brien a writer for fortune magazine eats dinner with What to Rent! owners Matthew Kuhlke and Adam Geitgey. The site adminsters a personality test to visitors and recommends DVDS based on the findings. As a more difficult attempt during dinner, the two will each pick someone in the restaurant and without ever talking to them, divine their single favorite movie. What to Rent! categorizes hundreds of films by star power, plot complexity, etc. Offering that individuals watch movie in the same manner that they interact with other individuals, in essence, relationships between individuals and movies are formed. Geitgey scans the restaurant and selects a guy delivering meals and cleaning dishes wearing tattered jeans, a steel bracelet with a few tattoos. He appears to be in his late 20s, working in a youth-trendy restaurant in the part of the city where people that age who don't have real jobs hang out. Geitgey offers, "Those are the kind of guys who barely made it through high school because they couldn't focus, but spend most of their time reading light philosophy books by singers-turned-writers like Nick Cave." He adds, "he would be interested in things that have an underlying philosophy but are also physically intense" concluding Starship Troopers fits that exactly! Kuhlke spots a waitress in her late teens or early 20s, black hair in a bob, very cute but she looks at the floor as she walks and avoids eye contact with customers. "She's unhappy...working around all these jerks who just want to have sex with her" adds Kuhlke. He also suggests she wasn't popular in high school and is shaking off her past by working in a cool place. Kuhle then states "She's like an old-school romantic comedy but she'd pick Breakfast at Tifany's, a totally crappy movie. She's cool enough to know she should pick Audrey Hepburn, but not cool enough to pick the right movie, Roman Holiday" so he battles between the two movies and finally decides on Girl, Interrupted! lol
According to many researchers "We don't just buy products, we bond with them. We have relationships with our things. DVD collections, iTunes playlists, cars, cell phones. Each is an extension of who we are (or want to be). We put ourselves on display through our purchases, wearing our personalities on our sleeves, literally and figuratively, for the world to see." And the web is transitioning from a search tool to a discovery one. The difference is that searching is what you do when you're looking for something, discovery is when something wonderful that you didn't know existed, or didn't know how to ask for finds you! "Everything you buy online says a little bit about you. And if all those bits get put into one big trove of data about you and your tastes? Marketers heaven.
To dig a little deeper, I actually went to What To Rent! to test our their application. It seems to be a content based recommender because it relies on an individuals personality and the content/factors of the movie to offer recommendations opposed to a rating system as seen in typical collaborative recommenders. The site uses the LaBarrie Theory, a movie viewer emotionally interacts with a film in the same manner that they interact with other human beings, to make suggestions. The site first decipher's a users' personality then forms a general model of how thety react to the world and their average emotional state by providing an upfront personality test. The cluser then compares your ideal stimulus and current mood to the possible relationships with movies on file. The film that maximizes the desired criteria is recommended. One of the major problems with personality tests is that users often lie (intentionally or unintentionally) about who they are and will answer questions as if they behaved in their ideal manner..leading to inaccurate responses. The system, instead, asks about seeminly unimportant life experiences an in attempt to get more accurate responses. After answering the 20 question personality test and 2 addt'l questions regarding my current movie and type of movie I wanted to watch, I was presented with their first recommendatino....Goodfellas a 1990 movie!! Definitely not what I had in mind. It's a movie that I have seen before and was hardly in the mood to see again. The next movie presented is Usual Suspects, this is in fact one of my favorites and I'm probably guilty of over-watching it at this point because Im not in the mood to see at right now either. In order, I was then presented: Raiders of the Lost Ark (1981), Kindergarten Cop (1990), The Shawshank Redemption (1994), and Jackie Brown (1997). Growing tired of their recommendations, I finally convinced myself that Jackie Brown would do. Less because I genuinely wanted to watch the movie, and more about me trying to find a recommendation that will suffice! While their overall stats boast a user satisfaction of 80.27%, 1 of 6 (more like 0.25/6) in my humble opinion is hardly accurate enough to warrant return visits to this website. I think I'll opt for word of mouth recommendations instead!
And by the way, the guy with the tattoos is the restaurant manager and before exclaiming Brianna Loves Jenna (a porn movie) he exclaims Starship Troopers and the waitress, as it turns out, loves Roman Holiday...perhaps they would do a better job of selecting movies based on individual photographs than relying on the personality test!
The race to create a 'smart' Google
What To Rent! (username: prs-test)
According to many researchers "We don't just buy products, we bond with them. We have relationships with our things. DVD collections, iTunes playlists, cars, cell phones. Each is an extension of who we are (or want to be). We put ourselves on display through our purchases, wearing our personalities on our sleeves, literally and figuratively, for the world to see." And the web is transitioning from a search tool to a discovery one. The difference is that searching is what you do when you're looking for something, discovery is when something wonderful that you didn't know existed, or didn't know how to ask for finds you! "Everything you buy online says a little bit about you. And if all those bits get put into one big trove of data about you and your tastes? Marketers heaven.
To dig a little deeper, I actually went to What To Rent! to test our their application. It seems to be a content based recommender because it relies on an individuals personality and the content/factors of the movie to offer recommendations opposed to a rating system as seen in typical collaborative recommenders. The site uses the LaBarrie Theory, a movie viewer emotionally interacts with a film in the same manner that they interact with other human beings, to make suggestions. The site first decipher's a users' personality then forms a general model of how thety react to the world and their average emotional state by providing an upfront personality test. The cluser then compares your ideal stimulus and current mood to the possible relationships with movies on file. The film that maximizes the desired criteria is recommended. One of the major problems with personality tests is that users often lie (intentionally or unintentionally) about who they are and will answer questions as if they behaved in their ideal manner..leading to inaccurate responses. The system, instead, asks about seeminly unimportant life experiences an in attempt to get more accurate responses. After answering the 20 question personality test and 2 addt'l questions regarding my current movie and type of movie I wanted to watch, I was presented with their first recommendatino....Goodfellas a 1990 movie!! Definitely not what I had in mind. It's a movie that I have seen before and was hardly in the mood to see again. The next movie presented is Usual Suspects, this is in fact one of my favorites and I'm probably guilty of over-watching it at this point because Im not in the mood to see at right now either. In order, I was then presented: Raiders of the Lost Ark (1981), Kindergarten Cop (1990), The Shawshank Redemption (1994), and Jackie Brown (1997). Growing tired of their recommendations, I finally convinced myself that Jackie Brown would do. Less because I genuinely wanted to watch the movie, and more about me trying to find a recommendation that will suffice! While their overall stats boast a user satisfaction of 80.27%, 1 of 6 (more like 0.25/6) in my humble opinion is hardly accurate enough to warrant return visits to this website. I think I'll opt for word of mouth recommendations instead!
And by the way, the guy with the tattoos is the restaurant manager and before exclaiming Brianna Loves Jenna (a porn movie) he exclaims Starship Troopers and the waitress, as it turns out, loves Roman Holiday...perhaps they would do a better job of selecting movies based on individual photographs than relying on the personality test!
The race to create a 'smart' Google
What To Rent! (username: prs-test)
Thursday, January 31, 2008
Search Inside the Music from Sun
Sun's Search Inside the Music
For those of us who love music, wishful thinking always includes the creation of a playlist that's specifically tailored to our interests and mood. Current recommender systems utilize one of two ways to sort music: collaborative filtering or content matching. Amazon uses collaborative filtering and offers suggestions based on what other's have found interesting. Pandora, as it turns out, uses content matching by hiring musicians to analyze songs and group them by hand into micro-genres. The Search Inside the Music poject from Sun aims to bring together the best of both worlds!!
Being the greedy individuals that we are, the best approach is one that offers the best of both worlds. Enter, supposedly, Sun's Search Inside the Music project! This project, led by principal investigator Paul Lamere, is is believed by team members to be the best music similarity algorithm because it's based on the actual sound. Leading recommenders categorize music based on artist, album, song and genre. Search Inside the Music analyzes features such as frequency, beats per minute, pitch, harmony, key, timbre, instrumentaion, tempo, intensity and energy level to map out the rhtyhm structure and determine the genre and which instruments are playing. In essence, it uses acoustic similarity to help people find music that "sounds similar" to music that they already like. According to Lamere, it takes a Pandora staffer about 20 minutes to categorize each song. The Sun system, on the other hand, is fully automated -- it's a computer-driven algorithm capable of whipping along at a clip of 3 seconds per CPU per song.
One of the more impressive capabilities of the Search Inside the Music system is its ability to generate a visualization of the acoustic distance among songs of different genres. This display shows which specific songs are similar to other songs of the same genre, but it also illustrates the degree of acoustic similarity between songs of different genres. This type of visualization provides a way to quickly create customized playlists based on acoustic similarity.
"For examples, let's say you've had a rough day at work; you're leaving the office and heading into heavy rush-hour traffic, and you want to hear myusic that will help you reduce your stress level as you drive home. The Search Inside the Music system can quickly generate a playlist that serves as a 'musical journey,' starting with higher-intensity songs that match your current stress level, and gradually diminshing the intesity of the songs as you make your way home. So during your 50-minute commute, you make the transition from Rage Against the Machine and led Zeppelin to Schumann's piano music--smoothly and seamlessly. And all the songs that are played are songs that you like."
Sun's other innovation is a tagging system that categorizes music based not on who's purchased it but on its attributes, described with tags like "quirky", "indie", "rock", "fast", "frenzied", "90's", or "cute" and "fun". For example, querying Sun's prototype search engine for Led Zeppelin brings up "tagomendations" such as the Rolling Stones and Jimi Hendrix. The user can then click a "why" button to find out why a particular song was recommended. Hendrix is recommended for Led Zeppelin based on tags like 'guitar gods', 'classic rock', 'guitar virtuoso' and 'psychedelic'. Sun is compiling these tags by searching reviews, lyrics, music blogs, social tagging sites and artist biographies, and incorporating the information into a prototype search engine. Compiling the tags based on a comprehensive search of the Web prevents people from gaming the system by generating their own tags to enhance the popularity of certain tracks.
In addition to recommendations for other music, the search engine provides links to videos, pictures and upcoming concerts, if the artist search for is alive and touring.
While there are certainly aspects/feelings that you can't get from analyzing the audio, it can help artists who are so new that they fly under the radar.
The more prevalent concerns about such a system stem from scalability. Where would analytical data be stored, how can the system be scaled to handle billions of songs and how can we find the computer power necessary to support such a system. Sun's computing model called grid computing, which lets you plug into a huge network of extremely power computers and draw on their combined CPU power on demand, provides a promising solution. Future advancements to the computerized analysis will enable recognition of major and minor chords, bridges and choruses, and the rhythm patterns of reggae, pop and ska according to Lamere. Though the Search Inside the Music project is not likely to be released as a commercial product, Sun officials say they plan to make the software available as open source, perhaps within six months.
A new music recommendation system from Sun (November 5, 2007)
Machine learning fuels Sun music recommendation technology (October 31, 2007)
Sun Micro spins its music software (April 9, 2007)
Search Inside the Music (September 26, 2006)
For those of us who love music, wishful thinking always includes the creation of a playlist that's specifically tailored to our interests and mood. Current recommender systems utilize one of two ways to sort music: collaborative filtering or content matching. Amazon uses collaborative filtering and offers suggestions based on what other's have found interesting. Pandora, as it turns out, uses content matching by hiring musicians to analyze songs and group them by hand into micro-genres. The Search Inside the Music poject from Sun aims to bring together the best of both worlds!!
Being the greedy individuals that we are, the best approach is one that offers the best of both worlds. Enter, supposedly, Sun's Search Inside the Music project! This project, led by principal investigator Paul Lamere, is is believed by team members to be the best music similarity algorithm because it's based on the actual sound. Leading recommenders categorize music based on artist, album, song and genre. Search Inside the Music analyzes features such as frequency, beats per minute, pitch, harmony, key, timbre, instrumentaion, tempo, intensity and energy level to map out the rhtyhm structure and determine the genre and which instruments are playing. In essence, it uses acoustic similarity to help people find music that "sounds similar" to music that they already like. According to Lamere, it takes a Pandora staffer about 20 minutes to categorize each song. The Sun system, on the other hand, is fully automated -- it's a computer-driven algorithm capable of whipping along at a clip of 3 seconds per CPU per song.
One of the more impressive capabilities of the Search Inside the Music system is its ability to generate a visualization of the acoustic distance among songs of different genres. This display shows which specific songs are similar to other songs of the same genre, but it also illustrates the degree of acoustic similarity between songs of different genres. This type of visualization provides a way to quickly create customized playlists based on acoustic similarity.
"For examples, let's say you've had a rough day at work; you're leaving the office and heading into heavy rush-hour traffic, and you want to hear myusic that will help you reduce your stress level as you drive home. The Search Inside the Music system can quickly generate a playlist that serves as a 'musical journey,' starting with higher-intensity songs that match your current stress level, and gradually diminshing the intesity of the songs as you make your way home. So during your 50-minute commute, you make the transition from Rage Against the Machine and led Zeppelin to Schumann's piano music--smoothly and seamlessly. And all the songs that are played are songs that you like."
Sun's other innovation is a tagging system that categorizes music based not on who's purchased it but on its attributes, described with tags like "quirky", "indie", "rock", "fast", "frenzied", "90's", or "cute" and "fun". For example, querying Sun's prototype search engine for Led Zeppelin brings up "tagomendations" such as the Rolling Stones and Jimi Hendrix. The user can then click a "why" button to find out why a particular song was recommended. Hendrix is recommended for Led Zeppelin based on tags like 'guitar gods', 'classic rock', 'guitar virtuoso' and 'psychedelic'. Sun is compiling these tags by searching reviews, lyrics, music blogs, social tagging sites and artist biographies, and incorporating the information into a prototype search engine. Compiling the tags based on a comprehensive search of the Web prevents people from gaming the system by generating their own tags to enhance the popularity of certain tracks.
In addition to recommendations for other music, the search engine provides links to videos, pictures and upcoming concerts, if the artist search for is alive and touring.
While there are certainly aspects/feelings that you can't get from analyzing the audio, it can help artists who are so new that they fly under the radar.
The more prevalent concerns about such a system stem from scalability. Where would analytical data be stored, how can the system be scaled to handle billions of songs and how can we find the computer power necessary to support such a system. Sun's computing model called grid computing, which lets you plug into a huge network of extremely power computers and draw on their combined CPU power on demand, provides a promising solution. Future advancements to the computerized analysis will enable recognition of major and minor chords, bridges and choruses, and the rhythm patterns of reggae, pop and ska according to Lamere. Though the Search Inside the Music project is not likely to be released as a commercial product, Sun officials say they plan to make the software available as open source, perhaps within six months.
A new music recommendation system from Sun (November 5, 2007)
Machine learning fuels Sun music recommendation technology (October 31, 2007)
Sun Micro spins its music software (April 9, 2007)
Search Inside the Music (September 26, 2006)
Wednesday, January 16, 2008
Emerging Recommendation Technology Helps Pick Shows
"Imagine you're listening to the radio in your bedroom. A song comes on that catches your ear - let's say "Ayo Technology," a hit single from the latest 50 Cent album. It's just the kind of jam that puts you in a good mood, but when it's over, a tinge of disappointment sets in. Surely the next song won't measure up. But what if it did?".........
Possessing an uncanny ability to relate to the sentiment above, I'm left to ponder what if the radio could predict what song I would enjoy hearing next....what a joy it would be to listen to all my favorite songs without ever having to skip a single one!!! And what if this same concept applied to movies, online games, and television shows alike! Apparently that is the direction recommendation technology is heading.
Recommendation Technology attempts to make educated guesses as to what else may interest you and is an extremely useful mechanism for any company selling items or any company that has a platform to distribute items. Companies have started realizing that perhaps the most effective marketing tool lies in their ability to personalize "suggestions" and tailor these recommendations to the interests of each individual. Given the plethora of information available via the internet, it can be challenging to find exactly what you want to find and as studies suggest, sometimes you don't know exactly what you're looking for which can amplify the problem.
While recommendation technology is not new, it has not been very effective in the past. As inferred in the "My TiVo thinks I'm gay" plot, perhaps recommender systems attempted to be too intuitive and less personalized initially. After recognizing the benefits available by offering more accurate suggestions (increased sales and session length) and the consequences seen by offering inaccurate suggestions (less sales, lower likelihood of returning), systems are seeking to become more sophisticated. "When you get down into the deep mathematical analyses of this stuff, and you have good data on content and users, you can make good predictions of what users will like."
Even more interesting is the fact that personalization isn't limited to just the web. Researchers believe that more than 1/3 of daily television viewing will be on-demand in the next 5 years. As an advocate of personalization, Im intrigued to see what other areas can be explored beyond the internet and television...
Emerging Recommendation Technology Helps Pick Shows
Possessing an uncanny ability to relate to the sentiment above, I'm left to ponder what if the radio could predict what song I would enjoy hearing next....what a joy it would be to listen to all my favorite songs without ever having to skip a single one!!! And what if this same concept applied to movies, online games, and television shows alike! Apparently that is the direction recommendation technology is heading.
Recommendation Technology attempts to make educated guesses as to what else may interest you and is an extremely useful mechanism for any company selling items or any company that has a platform to distribute items. Companies have started realizing that perhaps the most effective marketing tool lies in their ability to personalize "suggestions" and tailor these recommendations to the interests of each individual. Given the plethora of information available via the internet, it can be challenging to find exactly what you want to find and as studies suggest, sometimes you don't know exactly what you're looking for which can amplify the problem.
While recommendation technology is not new, it has not been very effective in the past. As inferred in the "My TiVo thinks I'm gay" plot, perhaps recommender systems attempted to be too intuitive and less personalized initially. After recognizing the benefits available by offering more accurate suggestions (increased sales and session length) and the consequences seen by offering inaccurate suggestions (less sales, lower likelihood of returning), systems are seeking to become more sophisticated. "When you get down into the deep mathematical analyses of this stuff, and you have good data on content and users, you can make good predictions of what users will like."
Even more interesting is the fact that personalization isn't limited to just the web. Researchers believe that more than 1/3 of daily television viewing will be on-demand in the next 5 years. As an advocate of personalization, Im intrigued to see what other areas can be explored beyond the internet and television...
Emerging Recommendation Technology Helps Pick Shows
Subscribe to:
Posts (Atom)