Tom Printy’s Blog

From mashable.com

Goodrec, a new social recommendation engine, has received $900,000 of a desired $1.05 million in funding from their series A round.  According to PEHub, the funding for this round included SoftTech VC and Chance Technologies Seed.

Little is known about the company to the general public as it is in locked testing for the moment.  What is known is that the service is looking to build a new system for remembering and sharing recommendations.  The project is the brainchild of Mihir Shah, a former product manager at Yahoo Search.  There is no word when the site will open to a wider testing, or even go public, but we’ll be sure to let you know.

I am not sure what “social recommendation” are, and how this differs from del.icio.us  or other bookmarking services.  I guess we will have to wait till they open up.

From alarm:clock

Israel’s Jinni has raised $1M in a first round from Start-Up Factory.

Jinni is developing a content search and discovery systems with personalization for consumers to discover movies, TV shows, and online videos. They plan to launch an affiliate program for Internet TV operators as well as to license its solution to global TV operators.

The premise is that consumers all gravitate to Iron Man and other blockbusters because they have a tough time discovering films and TV shows that suit their tastes. Yes and no on this. Consumers gravitate to IronMan because $200M was spent on it and because its nice to talk with others back at work who have seen the same movie. They can easily scan the film section to find other films that may or may not appeal to them but most don’t want to take chances on wasting a precious Friday night on a dud.

Jinni was founded by CEO Yosi Glick who was VP Marketing & BD at Orca Interactive.

There is a lot of competition in this space. It will be interesting to see if they are able to differentiate themselves.

Has either of these companies participated in the Netflix Challenge?  If they did and were able to score a decent RMSE they might have something.

I just finished watching “The Big Lebowski” on Hulu.  Hulu is the best new site to come out in a while. Were else can I get all this content for free and have it be legal? Sure there are some commercials but I can stand a 30 second interruption every now and then while I watch a free movie. There is also a ton of NBC and FOX network tv shows on the site as well.

As this service get bigger it would be nice if they implemented a recommendation engine but at least for now the movie content is small enough that I can figure out what to watch.

If you are looking for something to watch but don’t want to trek to the video store or wait for your next DVD in the mail you can always watch the content on Hulu.

Read/Write Web posted and article a couple weeks back about a study done by Two Wharton academics that asks “whether online recommendation services are a threat to the aggregate diversity of items discovered by their users. The study is titled “Blockbuster Culture’s Next Rise or Fall: The Impact of Recommender Systems on Sales Diversity” . The argument boils down to whether these systems help people discover only popular content or truely new content.

I believe that the point of a recommendation service to help retailers or marketers move inventory. People already know and desire the popular product (head of the tail), because it is in the social consciousness. There isn’t a need to recommend items that may have been just released. The manufacturer/publisher/content creater is already running some type of campaign to get a consumers attention. Instead, a recommender system should be putting items in front of a user that may have gone out of the social consciousness. Say for example a movie that came out 6 months ago in the theatre and is now being released on video, or a book that isn’t on some top 10 list anymore or even re-runs of tv shows. These are the items that you want to show users. They most certainly aren’t new but they may be new to the user/consumer.

There is a third part of the recommendation system that I think people and some companies are missing, and that is the conversation with the user.  Lets take movie recommendations as an example, when you start to talk to your friends about movies you sometimes say “You should see this movie it is great.”  After that initial sentence the conversation usually boils down to have you seen this movie. did you like it?  Subconsciously you are building a profile of your friend about what movies they like so that you can recommend them or not. Recommendation systems should do the same thing.  It should ask the users about movies that they may have seen or might like to see. People can’t remember all the movies that they have seen, but if you ask them about movies more than likely they have seen a good number of them.  This does a couple of things, first it helps the system build a more accurate profile of the user instead of relying on purchase history or rating random things. Secondly it gets the user involved, they like to give their opinions on things. Lastly it helps get ratings for those long tail items and helps with the proverbial “cold start” problem that one always hear about when talking about recommender systems.

If you want to see this at work check out the Sourcelight Discovery Guide. There you can see the conversation that is implemented. As you get more experience with the system it will attempt to go deeper into the long tail and have a conversation about those items. It also attempts to show you items that it thinks you have seen. So if you rate Indiana Jones it is going to give you all the Indian Jones movies to rate, because you most likely have already seen then and might enjoy rating them.

Are recommender systems a threat to the long tail?  Yes I believe that they are,  but they will never eliminate the long tail, just flatten it out some.  Perhaps when recommender systems get good enough we will have some type of double humped demand graph (camel tail) where initial demand is high then demand is high again in the future then it tails off.  People will always want to do what their social contacts do, initially they all want to consume the same things,  in private those seem people desire to consume individual things and often do. The job of a recommender system is to help people discover items to buy/rent helping your clients achieve their financial goals.

I learned about a new recommendation engine company today called mSpoke. They are based out of Pittsburgh PA and seems to have gotten their start in the financial markets, helping banker pick stocks. Currently they are deployed on two sites, their own rss reader called FeedHub, and website called Supply Chain Daily. Judging from these examples it seems that the mPower engine is tuned twords text based content, as both of the sites listed as example deployments are text based. I would like to see a job search engine based on this technology. It could send you tailored job listings instead of just keyword based listings that you get now from most job sites.

Minkey a new content based recommendation system surfaced today. Minkey uses a widget based approach to add personalized content to your website. From the looks of the widget it appears that they are targeting blogs and other content based websites. They claim that adding their widget will increase page views, create higher traffic, create more return visits, and of course create more revenue. The widget auto discovers content on your site and makes recommendation based upon that. I assume they are using some type of “click streaming” since there is not rating widget. Also it appears that they are using cookies to tie a user to their profile, perhaps annoying to some users who delete cookies often.

As well as offering the widget they are offering stastics based upon the widgets interaction with the users browser.

Adding the widget is easy. Create an account add your rss feed, customize your widget then insert the provided java script. I have installed this widget on the sidebar of the blog and I will watch it. This widget was much easier to install then the one by Loomia, however the Spotback widget has star ratings which I prefer. I will watch this widget and see what happens.

I learned about Adpinoin yesterday while surfing around. They are a recommender system for advertisements.  Essentially what happens is that users rate ads that they are show via a ratings widget on the ad itself. As you vote on ads the system builds up a profile for you, ads that you like remain visible, ads that you don’t like are remove and the system tells you “sorry”.  As you view and rate ads on the network it gets smarter about what ads to place in front of you based on what you and other like you vote on. From their overview page it appears they are using a clustering model for their collaborative filtering technique. This seems like a very nice way to model the user rating data.

This is really a neat idea and one that I have thinking about these past couple of weeks. I wonder if a better approach for gathering ratings on the ads would be using actual clicks on the ads as votes. If the user clicked on the ad then it is a thumbs up otherwise it is a thumbs down. Or perhaps use both, and have a graduated system where thumbs down = 0, no click = 1, click = 2, thumbs up = 3.  This might offer more data to the system and help build better clusters.  Also this would allow them to start gathering data on text ads instead of just banner ads that most people already ignore, or have their web browser filter out.

In any event I think this is a great first step and it will be interesting to see if these ads pop up on some of the websites that I frequent.

Today’s new recommender system is Spotback. According to thier website they are a recommender system that uses both collaborative filtering and aggregated knowledge technologies. They offer widgets that you use on your website or blogging platform. Once you add thier widget to your site you get instant Content recommendation, Rating tools, Top stories and Tag clouds. This is a very interesting approach to get adoption of your product.

They also offer a browser plugin that allows the user to rate items directly from the browser. I assume you are rating the whole page unless you are browsing flickr where their software seems to understand individual photos and that you want to rate only that photo.

If you are a user of their service you can login to the site and view your recommendations and what you have rated.

All of these services that they are offer makes this seem like they are trying to be many different things. This feels like a cross between MyBloglog + Del.icio.us + Loomia without any operant way to make money other than inserting ads into the recommendations. As you can see I have installed thier widget on my blog as well to see how it works. It was the easiest intall of a service like this to date since it was a native Wordpress plugin. You can see the rating widget below each post and a recommendation widget in the side bar.

According to Tech Crunch Loomia has landed a great cleint with The Wall Street Journal. Their widget shows stored that other people have read based on what you are reading. This is like Amazon’s people who bought this also bought this feature.

This is a nice deal for Loomia, it should help their exposure and having their “Powered By” link there really helps get the word out about them. Good Job Guys!

I really like the way Loomia works. They are a web service and have what appears to be three distinct products, a Free version a Media version and a Retail version. The free version is add supported, the media version is $50.00 per month. The reatail version is 2.5 percent referred sales. They are really covering the bases here with thier pricing models.

Setting up Loomia for you site is really easy. First you must publish feed of the items that you want recommendations for. Once you have done this you setup javascrip widgets in your site. These widgets interact with their service to produce content for you site. This is a very smart way to get rapid adoption of their service. This was so easy that I have set this up on my site here. I am using the clickstream widget with the recommendations showing on the left hand side of all my pages.

If Loomia really wants to get wide acceptance they will offer bloggers a rev-share with their free widget. A lot of bloggers are looking for a new way to make money from their blogs.

Seems like the recommender system space may heat up a little more this year so I copied my recommender systems review to its’ own page . I have also added the following review of Minematch.

Matchmine is a CA based company that uses demographic data to help with the “Cold Start” problem that many recommender systems face. They have built a flash application to help recommend movies (soon other categories). They would also like to partner with other recommender systems to collect data. I can only assume that what they are looking to do it collect market data that they can sell. I think that selling this data has great potential. “Data is king” and isn’t that what the new internet economy is all about selling data? It will be interesting to see if any other recommender systems partner with matchmine.

I have been watching the recommender systems space ever since I joined Sourcelight Technologies back in 2004. It is an interesting space that I don’t think many media outlets understand. So I thought I would put out some information about companies that I think are the major players in this space.

Sourcelight Technologies: This is the company that I have the most experience since I worked full time for them and still do on a consulting basis. The Company has been in business since 1997. It was started by a guy who owned a chain of video stores and some students and professors from Northwestern University in Evanston IL. Obviously I am biased towards the recommendation engine that Sourcelight produces.

Choicestream: This is the 500 pound gorrilla, based out of Cambridge MA. They have raised a ton of cash in the last couple of years. I beleive something like 35 million from at least 3 different rounds of funding. Currently they are being used by AO, Yahoo, SmartBargains.co, eMusic, Vongo, Akimbo, Movielink, DirectTV and Overstock.com. While this is an impressive list I am not ceratin that these are enough customers to satisfy thier investors. From dealing with them as a competitor the prices they are offering thier clients are WAY to low to support thier business. As of yet I am not convinced that thier recommendations are any better than any other approach.

Agent Arts (Fast Search): This company was based in Australia and now seems to be based in California. They make a system for recommendations as well as promotions. They primarily seem to be concentrated on mobile systems including Verizon, and infosapace mobile. I have never used thier technology so I cannot comment on how good or bad it is. I have spoken to Andrew Coates on the phone and through email. Seems like they are on a roll since they have been acquired.

Media Unbound: This is another group based out of Cambridge Ma. They have a bunch of people listen to music and record various aspects of music in order to recommend music. I have had Lunch with Michael Papish and he is a very easy going laid back guy who really gets music. So much so that he has his own record label, how cool is that? These guys also are little know to the media, however they power the recommendations that are used on MTV’s URGE digital music network. (Congrats guys). Thier music recommendations are super and I really would like to see a pandora like service from these guys.

Aggregate Knowledge: This company is realatively new and they claim to offer the first Collective Discovery Service. Not sure what this means really but if it means that you aggregate the data collected across multiple clients then this is in fact not a first and I am sure the all the other companies in this list do the same thing. This company is currently venture backed and has raised 25 million. As of yet I am not sure if they have any clients. Since there are no clients I cannot comment on how good thier recommendation engine is. I think it is intersting that they call thier product “Discovery Engine” copying Sourcelight Technologies.

Pandora (Savage Beast): Everyone’ favorite free online streaming system tried to seel thier recommendation system before making a go of it as Pandora. They follow Media Unbound’s model where they have employees listen to music and record attributes of that music. As far as I can tell they are no longer interested in selling thier recommendation technology but rather selling advertisements on thier site and in thier streams. They are going to seriously get hurt if they RIAA dosen’t relax thier royalty payments for interent broadcaster.

Criteo: The only company that I know of participating in the netflix prize contest. They are based out of France and seem to have picked up glowria as client. They supposedly have an api but the documentation for it seems to be down. They seem to have raised 3 million euros and have recently opened an office in California. Other then this I don’t know a whole lot about them.

Loomia : I actual thought that this company had gone away but it appears that they are back. They offer a total ASP solution with a free ad supported or pay per transaction solution. It seems thier only clients are some very low traffic web sites. These are the only guys who offer a free ad suported service. I actually tried this service on my old blog and it only ever recommended one story, Iturned it off.

Cleverset: This is another new company in this space. They are interesting in that you can sign up and start using thier service right away. Esentaily what they do is recommend products via javascript code on the checkout page of a website. This estentially follows Amazons user who bought this also bought….. They have clients like 1800backery.com and alpineer.com. From reading thier blog it seems like they are building the company to be aquired. I wish these guys luck.

I hope to keep this list up to date. If you have something to add please feel free to leave more inforation in the comments.