vanderwal.net

The Web, information use, and the failure of the spacial metaphor

Francis Cairncross' book title Death of Distance is a wonderful understanding of the world around us in many way and should now apply to spacial relationships on the Web. The idea of spacial relationships on the Web have been a stretch of the truth for a long time. Initially the idea of a person going out and "navigating" other spaces helped those new to the prospect of what the Web held grasp the Web concept.

The problem with spacial explanations of the Web is they do not work very well. The truth is we go nowhere on the Web, information is brought to us. The Web user is ego-centric and rightfully so, as the world of information and commerce on the Web revolves around the user. The Web is truly omnipresent. Information is everywhere at once. The Web can even follow the user on mobile devices. The user does not go out and explore different places, the artifacts of the places come to the user's screen based on what is of interest to the user.

I was reading David Weinberger's book Small Pieces Loosely Joined and it was painful to watch him twist and turn to get the spacial metaphor to work. A whole chapter in the book is devoted to Space [on the Web] (the book as a whole is very enjoyable and worth the time to read). Weinberger first discusses how we use the Web, using surf, browse, and go to a site. This is wrapped with an analogy explaining the Web is like a library where the user does not have access to the stacks of books, but a librarian (or clerk) goes and retrieves the book, based on the request the user made, and brings it to the user. He also states:

... this is perhaps the most significant change the Web brings to the world of documents: the Web has created a weird amalgam of documents and buildings. With normal paper documents, we read them, file them, throw them out, or sent them to someone else. We do not go to them. We don't visit them. Web documents are different. They're places on the Web. We go to them as we might go to the Washington Monument or the old Endicott Building. They're there, we're here, and if we want to see them, we've got to travel.

.... the odd thing is that, of course, we're not really going any place, and we know it.

This is just painful to follow. We keep bringing up this bloodied and battered spacial metaphor trying to make it work to explain more than the very tiny bit it did explain well. The spacial metaphor has long overstayed its welcome and it now hinders us as we try to build the future information interfaces, which include mobile information access and internationalization of information.

Yes, I am saying mobile information use is hindered by a spacial metaphor. It is more than hindered it is crippled by it. When prepare information now location is largely irrelevant, but access, device, application, and information form and highly relevant. Before we prepared information on paper and sent that information to people (which can be done today) and we largely knew how that information was going to be used. Today, with digital information the ease of information reuse and the user's ego-centric view of the information world, we must think of the user and how the information will be used. The proximity of the information to the user through access, storage, or personalization is what is paramount. Proximity is the only spacial element that has significance. This equally applies to internationalization as language and culture are the barriers to the information not space. A Brazilian may be sitting on the T in Boston and want to read the most recent information on rollout of WiFi in Rio. The user should not need to find the Brazilian neighborhood in Boston to get the information in the proper language (Portuguese) with familiar cultural inflections. The user can attract that information form easily, which can be brought to the user if that information and access have been prepared and enabled. The user may have come across a resource for this information while looking for a client's most recent press release and the user forwarded the link to her mobile device to read later. Access to information can and should be based on the users actions and choices.

The user can (and has been able to for some time) create their own metadata and retrieval structures. Communication with live people or machines that can and will convey useful information at the user's desire is not only the reality of the wired world, but what mobile use is all about. The user can set their proximity to information they have come across and connectivity conduits are enablers of that information they have yet to discover.

Up to this point the spacial metaphor only provided us with the navigation, but flat out failed us with what the user could do once they found what they were seeking. The user can browse, search, receive in e-mail (based on list subscriptions), read an information feed that brings to the user new information from sites the user likes or from aggregators, or a variety of other means. Once the user comes across information they have an interest in they want to keep that information attracted to themselves, via storage, putting it on a page that is accessible to a mobile or stationary device, and/or have the information delivered at a time that will be more convenient (getting a text message on your phone with the address and time of a party at an art gallery). Proximity also plays a role in location based services, such as bringing up restaurant listing and reviews when the GPS in our mobile device indicates we are near these establishments. The user should be able to identify favorites or preferences that can help provide "best options".

The realization of the failure of the navigation metaphor to provide for much other than a nice name for the grouped set of links that provide browsing options pushed me to investigate the Model of Attraction (MoA). The MoA is not perfect, but does provide a framework to think about information use and reuse as users currently interact with it. The MoA offers a method for us to work through how we allow the user to easily reuse information they found. The devices are just conduits for the attraction interaction to take place. MoA offers a framework that is also easy to understand, but is a literal description, which helps us see building, structuring, and preparing information and applications for the future.

Navigation -- R.I.P.

The future of UCD with attraction

Another snippet from Tanya from SXSW, this time from "Future of UCD" panel. Tanya picked out "users will not use item alone, but in a federation of devices", which is at the heart of the Model of Attraction (being presented at The IA Summit on Saturday March 22, 2003 right after the keynote). The future, which we are seeing pieces of now, gives more control to the user as to what they will do with the information and how the user wants to or will access the information. The body of research for Internet development has focussed too much time and effort on navigation (browsing is more encompassing or a term and more literal). Users not only browser for information, but search. The user is no longer constrained to a desk or building when they try to attract information they need to themselves and this difference greatly changes how we must think about providing solutions. It is long past time to retire navigations as a limiting metaphor and start working with a model that more closely represents what is literally happening. The navigation metaphor fails us as we try to encompass the future of information access, information use, and information reuse that has already begun to take hold around us.

One benefit of the Model of Attraction is that is provides a framework that includes information reuse. Many times an information application is built upon the perception that the output of the information form will be its only form. I have seen time and time again large organizations that have bought applications or built applications that only consider the initial output of information. That information form may be in a Flash movie, Acrobat PDF, PowerPoint presentation, Word document, dynamic Web site, or static HTML page (to name just a few options). What information creators do not consider is how the information will be reused. A PDF is great for printing or just reading, but pretty much fails for extracting information easily or having external pointers direct others to one piece of micro-content (a scentance, paragraph, or other delineated section). Each method of presentation of information has its own benefits and detractors. The one with the most legs is (X)HTML as is can be used on nearly all devices (desktop PCs, mobile handhelds, etc.) with little or no modifications, it is not the best medium for printing information, but if built to standards it can be easily converted and stored as the user desires. XML has the same promise, but one needs to work with a standard schema so that the information is widely useable and reusable.

Keep in mind the future is now. Our future needs metal models to help us build information applications and services for univeral usage.

Mobile use of Amazon Wish List and other adventures with mobile gadgets

Tonight was a mobile/portable device adventure. Joy and I had a wonderful dinner at Thyme Station in Bethesda. A Japanese couple sat down next to us and we could see that they were having trouble with the menu. The guy has a tiny Sony device, about the size of an Altoids tin, but a third of the height and finished like an Apple TiBook. He was using the translator and not finding what he was looking for and Joy nudged me and asked if I thought if we should help. Joy asked the couple if they needed a little help the couple giggled and then nodded. The seemed to be having the most problem with the sirloin kabob and did not know what it was. We explained it was beef cooked on a stick. They returned to tapping in word and getting translations on their device and passing it back and forth. They also seemed to have some entertainment on the device. I really am kicking myself for not getting a picture and talking with them a little more.

Joy and I tried to figure if we could catch a movie and checked the movies and listings from the Hiptop. The couple next to us seemed equally interested. We were had just missed the start times for the movies and were going to have to wait for the next round at 10 or so.

We settled on going to Barnes and Noble instead. I found a book I had read about in some RSS feed this past week and really liked it. I was not ready to spring for the book at full price, so I pulled up Amazon and checked the price on the book (Stop Stealing Sheep and Find Out how Type Works). I found the book and a 30 percent discount. But the best part is I dropped it in my Amazon Wish List, which is where I keep track of such things. I then check other items in my Wish List to see if they were in stock at Barnes and Noble. I could not find the two books I really have been wanting and could be willing to pay full price, which means I may buy off of Amazon with a couple other items that will bring me enjoyment.

My Amazon Wish List is one information store that I like having at my side. This really highlights the element of the Model of Attraction that focusses on the user being able to have information be attracted to themselves and have this rough cloud of information follow the user for use when needed. The user works to find information they are attracted to, by searching or reading articles or blogs that the user has an affinity for. Once the information is found most metaphors and models stop working (navigation, information foraging, etc.) but the MoA keeps on working. If the user has a strong enough attraction she may want a method to store the information for further access. How and where become the next questions. The next step is accessing from storage (or your personal information cloud that follows the user). Mobile devices are tools that allow for this attraction to continue, but how do we ease this use? Personalization adds another layer to the user setting attraction (like Amazon's wish list).

Information Accessing and Browsing book

I am really enjoying Accessing and Browsing Information and Communication by Rice, McReadie, and Chang from MIT Press. I am less than 100 pages into this academic review and assessment of how people approach, find, assess, and retain information. The book is wonderful as it takes an cross section of academic disciplines and the research from them on browsing information and information use. The book is uncovering new pools of information in the library sciences, organizational communication, and knowledge management fields that help one better understand how people find and process information. There is a lot of information that is echoing the Model of Attraction approach to user and information relationships as well as helping to better refine the approach (there are some updates to the MoA that I have clarified in the last three weeks, and this reading is helping gell).

I may write a full review when I finish the book. I am reading the book on the train to and from work at the moment. I may need to read at home so I can check information that is being referenced.

Mobile phones offer better control of communication

The Beeb offers Mobiles let you control your life. Increasingly the mobile phone has given people the ability to control their communication flow and their information streams. I can relate to this as I use my mobile as my primary phone, actually I have two mobiles and one is my mobile information access point. I use the my voice mobile for talking with friends, family, and work, but it works wonderfully as it does a much better job with voicemail than my work phone. I use the other phone, the Hiptop for mobile e-mail to have needed information at my fingers and to send notes to my self that are easily used from e-mail. The Hiptop also offers the ability to touch base with coworkers and digital friends as well as keep up with reading off the Web. I agree with the article I have better control of my communications with these mobile devices. I am also getting closer to having information attracted to me so that the information is there when I need it.

Understanding Referrer Systems

Rashmi Sinha's Recommender Systems overview on Sig-IA is a great overview of the tools used to bring information closer to those that may have an interest in that information. Essentially this gets to the heart of the Model of Attraction. Recommender Systems can provide tools that group information through links that are based on others similar interests. Finding information is what many spend much time doing. Lessening this time and providing methods to predict interests are helpful. This of course takes patterns. Patterns need a breadth of data to be useable.

In particular you may be interested in Rashmi Sinha's "Interaction Design for Recommender Systems". [hat tip Matt]

Hiptop helps show extending the model of attraction

I am already enjoying my Hiptop for much of the reason that I picked it up. I wanted access to information. More importantly I wanted information to be able to follow me. I found information or thought of information I really have been wanting to have access to that information from where ever I am. I wanted the ability to share the information from where I was and have others be able to use that information to better their understanding.

Yes, I have had cellphones and have called others, but the information is not that useable in voice form. The information needed to be convered to data elements that could easily be used and reused. Voice only (at the current time) allows us to hear then act upon the information and not store that information in a searchable repository or to easily share that information back out.

Yes, I have PDAs (Palm-based handhelds), but they need to synch with other devices to share information and the e-mail capabilities were not the best around. The 3rd party applications on the Palm and the fantastic operating system that is fast and small are great features that will be hard to beat by anybody.

I have been looking for a solution to have the information I wanted when I want or need it in my hands. The Hiptop gets me much closer to that goal. I tend to use e-mail to share ideas with myself and others. This weblog is another method of doing the same. Being able to search for an address and get a map is a solid tool to have at all times.

This is a personal quest to have the Model of Attraction (MoA) extend back to myself. The MoA not only helps us think about the attration between the user and information during the finding tasks, to help improve findability, but in phase where the user wants information to stay attracted to them. My Hiptop is my information attraction device. I can push an e-mail to myself that has the name, address, time, and phone number needed to do to a party with friends that have come in from out of town. I can access my Amazon Wishlist when I am in a store to help remember the author or title of a book, CD, or DVD I have been seeking. This bookstore amnesia (or musicstore amnesia) can be a thing of the past. The Hiptop provides me the information in my hand and gives me the access to the information I do not have at hand wirelessly.

There will be some experiments to see if I can improve on the information attraction to keep the information closer to me. Am I getting rid of my Palm? No, as there is information in it that I prefer in the format it is in. I will be keeping my cell phone as it has great reception and is CDMA (I found having a non-dominant cell phone technology is an advantage during emergency times, like being in San Francisco during September 11, 2001, which is a TDMA and GSM dominant city. I was one of a few that had no problem getting a signal to call out). It is rather awkward having three devices with through out the day. We will see how it goes.

Model of Attraction moves forward

The upcoming IA Summit in Portland, Oregon is providing me the opportunity to offer the Model of Attraction live and in person. In the coming weeks I will be posting background to this presentation in digestible chunks. You are free to peruse the initial draft of Model of Attraction from March 2002, the Model of Attraction outline from December 2002, and the attraction category here in Off the Top.

Navigation is Broken

Part of the need for developing the Model of Attraction revolves around the broken metaphor of navigation that many IA's put much trust in. Metaphors use a concept that is understood (often not related to the topic at hand) to describe the hard to understand or the new. All metaphors limit understanding as they do not accurately describe the actions and relationships, they only provide a framework that helps understand bits of the whole. The navigation metaphor has been stretched beyond its limits and has limited the possibilities of information structures. We as IAs are worse off by leaning on navigation beyond its narrow boundaries and the users of the sites and the information bound in the sites are worse off by the over reliance on the navigation metaphor.

To see where this discovery began, go to this discussion on IA, navigation, and information space. Pay attention to Stewart Butterfield's comments. In addition navigation does not permit us to think about understanding visual attraction, reuse of information, information access methods outside the PC based Web, mobile access, personalization, content management, or the ability to have a rough cloud of information that follows the user (access to information where and when one needs it). In the Peterme discussion I stated I would sleep on a solution. I repeatedly slept and woke thinking about this problem and fell into the Model of Attraction early last March and have been working with the MoA since then.

I have used the MoA with clients and when mentoring IAs and Web developers. The comprehension is much better when describing the same approach than when using the navigation metaphor. Clients quickly understand the need for controlled vocabularies based on the user's common language and understanding. MoA helps easily explain the need for grouping of information around categories and facets. Card sorting tasks become easy to understand for the clients and helps them assist in the process. Most of the IAs took kit (persona, taxonomy, wireframes, metadata, etc.) are more usable as their need is easily understood by all in the context of attraction.

Scent not Strong

You may be thinking this rough explaination you are getting sounds like Information Scent. You would be right, to some degree. But information scent, like navigation, is a metaphor. Yes, scent breaks too and is quite difficult to use with clients as some things get very confusing for them. Scent helps IAs understand what is going on a little better and there is great research that has come out of the Information Scent community. But when you get down to it scent is a subset of attraction. Scent is one method of attracting or repelling the user toward information they are seeking and keeping extraneous information out of the mix. Scent also has its limits. The scent metaphor works with getting the user to information, but it gets very murky when the information needs metadata (to help augment the attraction between the user and the information they are seeking). Scent and odors have distinct understandings and altering the scents for understanding (metadata) raises many questions from clients as the client tries to understand. Scent does not work well when trying to build information structures for mobile access to information, nor for setting the ability to have the information follow the user (What you want to use a blueberry muffin as a perfume? Don't think so).

More to Come

This only defines MoA by showing the limits of navigation and scent. More understanding is on its way in upcoming weeks and will be put in a presentation format for the IA Summit. Those that are worried, we are not throwing out navigation or scent. We are keeping navigation in its small usable space where it works well. Scent has provided great research and has similar properties to attraction as it is a subset. The Model of Attraction will provide a broader foundation that allows us to move into the future as we build information structures that include possibilities for mobile access, social networks, and true access to information as the user needs it by keeping the information close at hand. The MoA does not solve these problems, but provides a framework that does not break when we work to solve these issues.

Trust in the Friendster and Ryze social networks

I finally got around to joining Ryze a social networking site. After just a few hours I find that I am more impressed with Friendster for many reasons.

Part of my preference of Friendster is the trust factor, which Friendster takes advantage of and Ryze has left off. When I go to add a person to my friend list in Friendster, Friendster sends a note to the person asking permission. Upon approval from the person I wish to link as my friend the person shows up in my friend list. Friendster also provides "introductions" through people on your friends list and people can recommend friends to others. The testimonies on Friendster are a valuable commodity as they must be approved by the person receiving the testimony (granted this can become rigged, but it offers the ability to give a better sense of the person). Friendster also shows the relationship to others in the whole network. These are assets that are very important to me when building trust and components that are missing in Ryze

The focus of the two systems is different. Ryze purports to be a business networking tool, while Friendster is a social networking tool. The lack of the above mentioned features for trust in the Ryze tool seem to make it more vulnerable for abuse or misuse. In a business context trust is very important as investment and decision that can have a determinant impact on success are made from information. People are one conduit of an information flow.

To me, it seems important to have the ability to have a some measure of trust built into a representation of a human relationship that then provides information the user may then act upon. It is also important to know that this relationship is approved or reciprocal People entering my page on Ryze from a Google search may see that I have Peter Morville as a friend. The Ryze tool does not provide a mechanism to verify this is the case. The only method one may have to determine if Peter and I are friends and there is a mutual trust that may be drawn from that relationship is to see if Peter has listed me as one of his friends. The Friendster model asks the person being listed as a friend if they agree to be listed as a friend. This model also makes it easier to find connections between people.

The Friendster tool can help people find others for various reasons or used as a tool to open doors to new music, movies, or interests. In the week or two I have been on Friendster I have found new music to dive into. I have been bugged by two folks to add my music preferences as they are looking for new music to explore. This would also be helpful for Ryze in a business context, but I don't see digital trust mechanisms that would ease this transaction. Ryze is more tool to provide links and correlations for serendipity then a quick way to find people of similar mind sets or people to expand intellectual and/or monetary pursuits.

I will stay with both social networking tools, but I know which tool I have more trust in at the moment. The trust factor is built over time through interactions that ring true to one's own meter or metric for judging. Friendster provides a step up on the trust meter. The best analogy is a sidewalk meeting: Ryze provides a glimpse of two people walking near each other (we do not know the relationship as it could be victim and stalker, friends, or two people that started chatting on the subway); Friendster classifies the relationships of the two or more on the sidewalk and shows the approved relationship as friends.

Peel exposes layered storytelling

Design Interact examines the Seattle design firm Peel and their layered storytelling approach to information structures. Layered storytelling is explained:

Layered storytelling means that a site opens much like a film, with a splash of music, photography and animation, but not a lot of information. If you stay on the top level of the site, your experience is similar to watching a documentary on television. But if you click on any topic, you dive down into a more book-like experience, with long texts and additional background information. The idea is that a visitor skims along the surface until he or she finds something interesting and then digs in to read more.

This appoach provides the ability to have a one way interaction with the site as it entertains and informs, but when the user is attracted to a topic, idea, or visual cue they can interact and find out more. I have enjoyed the layered storytelling approach when I have encountered it. It does seem like it would have the same repeat user problems that other multi-media interfaces encounter, in that having to wait for load times before interacting or navigating is usually problematic. Providing an option to use the layered storytelling or providing it the first time by default (but if a user is like me and works with three or four browsers open or working from many computers, setting a cookie to track repeat use will not solve the issue).

This too is worth coming back to as it provides intamacy with the user and a topic. This can help break down some of the dry appearance of some dull topics that are difficult to unwrap, like sciences, urban planning, the history of duct tape, etc.

Emergance finally makes my reading list

My other reading on my quick trip to Spokane, Washington included Stephen Johnson's Emergence, which I am finally getting around to. It is a wonderful book that cuts across many fields of expertise and ties them together in a well thought through manner. Not much in the book is really new, but the connections of the cross-currents makes a fun read. It has sparked the Alan Turing interests in me again and has me looking for my Metamagical Themas by Douglas Hofstadter, which must still be in a box. Metamagical opened many of the doors Johnson opens in Emergence, but is a more approachable manner. I will hopefully finish Emergence on my next quick jaunt.

Model of Attraction Outline - Version 1

The Model of Attraction ouline version 1 is now posted. The outline has been structured to set up a structure for filling in the blanks and providing a better strucutre for understanding the MoA. Outlines are my foundations for writing more serious works. Outlines help me find holes and provide a structure to rest content upon. This verion is largely attributed a train ride to Philly that allowed me time and untethered space to think, order, and write.

Please comment if you are so inclined. Find holes are areas that do not seem fully fleshed out enough. Thank you in advance.

Findability explained

Peter Morville finally puts his findability explanation in writing for all to see (in the wonderful site called Boxes and Arrows). The idea of the term and meaning of findability is growing on me. Findability is a solid lead into the problems of information structure. The explanation of how to start fixing the problems and actions needed to help eradicate the problem can reside in the method/model of attraction (an update to the MOA should be available in two or three weeks, extenuating circumstances have slowed the updates and progress).

I have updated Eric Scheid's initial entry of the Metaphor of Attraction at IAWiki. I will try to keep that up to date as well as the MoA information here. I am preparing an update to the information, but as there is an out of town wedding this weekend and other diversions this next week it may be a short time longer.

I am looking for a better name for the metaphor of attraction. I am thinking about: Principles of Attraction; Magnetic Informaiton; Information Attraction; Attractability; or something else. Please provide feedback if you have ideas that would help. Thanks.