Loading...
Flash Player 9 (or above) is needed to view slideshows. We have detected that you do not have it on your computer.To install it, go here
- All Comments (0)
- Notes on Slide 1
- HiAnd thanks to Eric for inviting me to speak at Oz-IA 2008.I should probably add the word search to the title of my talk because I plan to delve a little into this as well.
- I don’t know how many of you saw the recent Doctor Who episode “Silence in the Library”?It was an interesting story about a whole planet that has been transformed into the biggest library in the universe.Every book ever written, anywhere! However, when the Doctor and Donna arrive the place is empty, but we soon discover that at the last count there were over 4,000 people in the library.All this information, but no one is able to access it.
- The answer lies in the shadows: Swarms of microscopic flesh eating things, which formally lived in the forests of the planet that were destroyed in order to produce all the books.It is as though the process of accumulating all the information held within it the seeds of its own lack of usefulness. This got me thinking about the web, a wonderfully vast resource of information, but a resource on which we are increasingly dependent; - maybe because of the convenience, - or habit, - or perhaps as a result of marketing and our growing addiction to speed and instant gratification in a consumer-driven world.But, do dangers also lurk in the shadows of the web?
- I use the web all the time and I use it to search for information. My daughter and her friends virtually depend on the web and search engines for all their school assignments. And yet, there are now comments from academics around the world about the decline in basic research skills. A university lecturer recently bemoaned the fact that nearly all the student assignments he marks now cite the same few resources (mostly from the first page of Google results) and very rarely go back to primary sources.In a survey of over 2,000 faculty members from different universities in the US, nearly a half said they felt the internet was having a poor effect on the quality of student work.Given the growing reliance on online information, I find it curious that many clients seem to be willing to hand the process of helping people locate and retrieve information from their sites over to other people – either Google or the folk.
- \'Most people now use Google, so site navigation isn't that important anymore,\' Replied a recent client when concerns were raised about the proposed Information Architecture.How many have heard something like this?When it comes to Information Architecture, we can find ourselves facing two opposing forces:On the one hand clients, who basically feel the answer to finding information on their site is best left to Google. And, so in some cases are willing to spend a fortune on search engine optimisation but are unable to find any money for user testing.
- The other force is developers. Often highly skilled people who are up-to-date with the latest trends and understandably want to push the boundaries of their craft.Embracing with unbridled enthusiasm everything and anything new.Or perhaps they are enticed by the tyranny of the cool.
- A desire to out cool the coolest, or even more important avoid looking “uncool”.“Coolness” can be a very transient thing!I remember when the coolest were those who could exactly replicate a printed page with pixel prefect layout and hundreds of spacer gifs.And what about Frames, how cool were they back in the days of Netscape 4? My first commercial site used frames and that was so cool it was positively frozen! Mind you, some people had problems using it, but what they heck they were uncool!But I guess today most of us view frames much like we would a piece of over-cooked limp cabbage: Tepid and certainly not cool.Where as social networking and folksonomy – now that’s cool.
- To some extent, the move to hand everything over Google and the drive to incorporate a more social-networking approach are contradictory.The growing reliance on search results could cause web users to become increasingly disengaged from the process of exploring for information; whereas, the whole success of the social software movement is predicated on greater engagement in the process by an ever increasing number of people. I am not about to say one approach is better of worse than the other.Rather, I wish to raise some issues that you might like to balance up when trying to find the approach that best suits:The users of your siteThe interests of the site and its proprietor
- It is easier for any discussion about the internet to degenerate into the good guys – bad guy paradigmPC vs AppleMicrosoft vs GoogleProprietary vs Open sourceWith one side painted as the devil and the other as angels.In my view, Apple, Microsoft and Google are all just companies out to get market share and return a profit. None are either better or worse than the others.When talking about search I will be largely talking about Google. But I am setting out to neither praise or diss Google – just use the company to illustrate the points I wish to raise.
- The Pew Internet and Lifestyle research results for last month confirmed the increasing use of search engines in the US with just over 49% of internet users using search daily. The Web holds billions of pieces of information and so good search tools are essential for helping us to locate what we want in this vast sea of data.And Google, with nearly a 70% share of the search engine market, is a good search tool: It must be for it can return millions of results for most search requests in a fraction of a second.
- But, is a million results what we are after? For most people, Relevance is the key to determining the usefulness of all these results.We all know how frustrating it can be to troll through search results that don’t match our expectations.And relevance is largely determined by two factorsRecall: That the percentage of relevant documents retrieved from all possible documents. Precision: Which is a measure of the percentage of documents retrieved that actually match the searchers needs. As both site makers and users what we are after is Full recall with high precision – This occurs when all the relevant documents that match our needs are found. The big question is, who decides what is relevant?
- We have probably all heard at sometime or other the idea that the British mathematician and scientist Sir Isaac Newtown discovered gravity
- And a quick Google search using “discover gravity” brings up five and half million results with at least three on the first page seeming to confirm the idea that Newtown did discover gravity.One of the results links to an educational resource – TEACHER TUBE
- With an entertaining animation showing the kiddies how an apple fell out of a tree and when it hit Newtown on the head, he said “aha, gravity.”A sort of bath full of water Eureka moment!Now clearly Newtown wasn’t the first human to notice things fell down!Building on the work of Copernicus, Kepler and of course Galileo, who had all, surprise- surprise, also observed objects falling, Newtown worked out the mathematics and developed the law of Universal Gravitation to explain what was happening.Now this served us fine for the next 300 years, but eventually it was found that Newtown’s law wasn’t exactly universal and Einstein filled in a few holes with his theory of relativity. So what is the point of this?
- As Professor Koehn from Harvard Business schools puts it, facts alone are not knowledge and great volumes of data don’t necessarily improve our lives.Frank Zappa said much the same in a more strident fashion in the 70’s avant-garde rock saga Joe’s Garage.Facts and information only become useful knowledge when they are in context and their accuracy can be relied upon. But this can take time, and in our fast consumption-orientated world we never seem to have enough time.
- Benjamin Barber in his book “Consumed” describes how societies like ours are now controlled by consumer capitalism: Our basic needs have all been met, so goods are no longer produced to meet needs, Rather, needs are manufactured to stimulate the marketing of goods.An integral part of this marketing effort involves the infantilization of adults which has at its core three dualisms:EASY over HARDSIMPLE over COMPLEXFAST over SLOWA form of compulsory attention disorder in Barber’s view, where we are increasingly driven by the child like imperative for instant action and reward.
- But back to search: I don’t want to suggest that there is anything wrong with searching the web. But I do believe there is a potential for problems to arise when it becomes the only way of obtaining information.I should make it clear at this stage, that my observations don’t relate to internal site search facilities. Generally speaking well made sites will incorporate an effective search facility as one part of the overall information retrieval process.What I am concerned about is the increasing, and automatic, use of web-wide search and the growing acceptance that this will best serve users and sites
- At a very basic level, the growing reliance on search taps into the infantilization of adults described by Barber, where the desire for instant gratification, which was one considered the hallmark of a troublesome two-year old, is now acceptable adult behaviour.I want the answer and I want it now,Preferably without doing too much work.Search seems the logical way to go – Quick and Easy.Bang in a request, get back a squillion answers.But for me, this raise a number of interesting questions:How useful are the answers?How many pages will I have to look at to find the one I want?In short, who or what determines which answers will be one page 1 and which will be on page 253!
- No one outside Google knows exactly how Google works, but we can get some insight into the initial thinking from the research article The Anatomy of a Large-Scale Hypertextual Web Search Engine written by the Google pioneers Sergey Brin and Lawrence PageOne of the big ideas behind the success of Google was the introduction of “page ranking”.And, if you will excuse another QUOTE from this article“A page can have a high PageRank if there are many pages that point to it, or if there are some pages that point to it with a high PageRank. Intuitively, pages that are well cited from many places around the web are worth looking at. “END QUOTEAnd boy, did page ranking work or not? Within a few years Google was synonymous with search
- Following Google’s outstanding growth in popularity, most search engines now use PageRank or something similar to help determine the order of search results.In general, PageRank considers a page important, or of high quality, if it is linked to by many other pages on the web. As a result, these “currently popular” pages routinely become the top search results, and so in turn become even more popularIn a very interesting research project, Cho and Roy, from the Computer Science Department at UCLA considered the impact of search engines on page popularity. They proposed several theoretical models and studied the impact of page ranking for about 5 million pages from 150 sites over a 7 month period in 2003.They found that when search engines rank pages based on popularity, it can take several orders of magnitude more time for a new page to become popular, when compared to the pre-page rank world. This even applied to pages of high quality.
- Since Cho and Roy’s work, Google have refined the way they rank pages and search results, but the process still tends to favour the popular over the new or esoteric. Hardly surprising considering what are the most popular search requests:Over 90% of the top ten results for the 5 years 2001 to 2005 relate to popular culture:Britney Spears made it every yearBrad and Angelina made three appearances as did Harry PotterHowever, Osma bin Laden only appeared once in 2001Irag in 2004 and Hurricane Katrina in 2005While search is probably a great way to tap into the popular culture of the time with simple search requests, It is not so good at handling abstract ideas or for finding less “popular” information.Combined with the problem of getting new and “less” popular pages noticed, is another problem well recognised by the Google founders Brin and Page:QUOTE“The number of documents in the indices has been increasing by many orders of magnitude, but the user's ability to look at documents has not. People are still only willing to look at the first few tens of results.”END QUOTE
- Most users do only look at the first page of results, with less than half going beyond page 2. So if web-wide search is the prime way people are going to find something on your site, getting on the first page of results is vitally important.So important in fact that it has stimulated the huge SEO industry and great profits for Google.Now of course Google doesn’t run ads – it has “sponsored links”And this raises another, interesting question: how many average web users in Australia recognise the difference between the normal search results and the “sponsored links”?Google say the difference is obvious.The ACCC however, took a contrary view last year, saying Google was engaged in misleading and deceptive conduct in the way it presented “sponsored links” in the results for some search requests.And sought an injunction restraining Google from publishing search results that \'do not expressly distinguish advertisements from organic search results\'.
- Now while I personally don’t agree with the ACCC point of view on this, it brings me back to my original question about why are clients willing to hand the process of locating and retrieving information from their sites over to the vagaries of search engines.It seems to me that this can only have two possible long term outcomes:First, it could diminish the brand of the site, for I would imagine most site owners would prefer their brand, not Google, to be the first thought that springs to mind when some one is looking for the information or services they provide. This I feel would be particularly so for Government agencies charged with the responsibility of serving the community.Second, it is likely to incur increasing ongoing costs for site owners as they struggle to maintain a presence on the first results page.I am not decrying the value of S.E.O., just that it shouldn’t be done at the expense of developing and testing an effective site architecture.In Jakob Nielsen’s view this is going to become a very important issue and the money spent on user testing offers a better return on investment than that spent on paying for search engine ads.
- Few years back, the notion of traditional site taxonomy and Information Architecture came under sustained attack.There was a view, that as the web was becoming more interactive with greater amounts of user-generated content, the “top-down” approach of site owners deciding where things were on sites was no longer relevant or effective.Sadly, any discourse about the roles taxonomy or social networking have to play in the classification of content quickly degenerated into a slanging match. For some reason or other, it just seems that when ideas are discussed by the web community, common civility often flies out the window!
- Social software seems to encompass a “bottom-up” process of development, where it is the users and contributors, that is the community, who define actions and control outcomes, rather than some external entity.This can manifest in many ways such as blogging, social networking sites like Facebook, or sharing sites like Youtube.But, when it comes to discussion about the role social software can play in the classification and retrieval of information from large sites, I am mainly interested in the potential of tags and tagging.Arguably, tagging gives users, that is the folk, greater say over how content is classified and retrieved. And, Thomas Vander Wal coined the term folksonomy to describe this process.
- Folksonomy is about allowing users to categorise online content that interests them, by using words and terms that they are familiar with. But, the tagging of content alone does not make for a folksonomy, since it is in effect a private system of identifying stuff. It is the sharing of these tags by different users that underpins the notion of a folksonomy.When it comes to information retrieval, it appears that a key assumption is that, if enough people tag a web page or some object, interesting and useful patterns will eventually emerge.But it seems to me that this assumption is based on three presumptions, namely:A significant number of people know what tags areThe overall quality or usefulness of the tags providedAnd, a willingness by people to use tags when trying to locate resources on a siteIn order to test these presumptions I undertook a small research project in 2006. I won’t detail it here, as I know some of you have already heard me talk about it at other events – however I would like to take a few moments to briefly summarise a few things
- The research project involved a quick survey of forty people, in four groups of ten and I was interested in getting answers to three key questions:How aware were they of tags and social bookmarkingHow would they tag somethingAnd, how likely were they to tag in the future.At the commencement of each test sessions, the participants were asked if they had tagged any web content in the past. And 8 out of 40 said they had. They were then shown a photo on Flickr and the tags associated with itAnd given a brief description of tagging
- Each participant was then asked to tag:This photo of Sydney Harbour at sunsetAnd this photo of a red back spider with its prey.They were told they could provide as few or as many tags as they liked for the photos.The participants provided a total of 116 tags for these two simple photos. 76 of these tags were unique, that is they were provided by only one person. Of interest was the relatively low use of what I felt would be common or important tags such as: Harbour bridge, Australia and New South Wales for the photo of the Harbour bridge. And, biology, dangerous and poisonous spider for the red back.Also the idiosyncratic nature of some to the unique tags: ‘Manly ferry’ and ‘yachts’ when neither are in the picture of the harbour. Or, Sheds, green and slim dusty for the spider.
- At the end of the survey, the participants were asked to indicate the likelihood of them providing tags for web content at sometime in the future.Most indicated a reluctance to tag, but just over half said they would do so; sometimes, often or always.Here are a few of the participant commentsIn general, they felt tagging would bring them little personal benefit but would be an altruistic thing to doAnd, most felt they wouldn’t have enough time for such altruism.
- At the beginning of this talk, I made mention of the pressures that are sometimes brought to bear on a project by the enthusiasm of web professionals keen push the boundaries and expand the ways we use the web.I am as guilty as the next person of this tendency to be blinded by enthusiasm. A while ago I suggested to a client with a natural history-cultural site, that it might be a good idea to allow members to put photos on the site – the reasoning being they had a lot of good things to photograph and people like taking and sharing photos, so why not.Well, they put a very nice members photo display on the home page and seeded it with 4 photos and some months later there are still only 4 photos!
- The Brooklyn Museum in the US on the other hand, is very community orientated and has heaps of user photos. It also allows people to do all sorts of other things on their site. In the Blogger area for example, Museum staff write about what they are doing and site visitors can comment on this work. In the last year, there have been 7 postings by staff, many of them extremely interesting and yet they only appear to have attracted a total of 11 comments. Five relate to a fascinating detective story about an Egyptian mummy.
- Another posting about a photographic exhibition called “Click” appears to have no comments, whereas in fact there is a fair amount of discussion, only you have some understanding of what “tagged” means and then know that that you need to click the word “Click” either at the end of the article or in a tag cloud in order to find them.All this got me to thinking once again about how much normal people, that is people who have a life outside the web, use these new features we find so enticing. After several unsuccessful attempts to find an answer with Google, I decided to do another survey.
- For this quick survey my basic aim was two fold:First to gain some insights into how much people use some of the web 2.0 type of stuff that is now available. And, second to compare the extent to which people who work making web sites and those who just use sites, use these tools
- I prepared a simple questionnaire which considered these issues:How many people provide comments on web pages of othersHow many people have their own blogs and comment on the blogs of othersHow much people had tagged content and how often they used tags and tag clouds when retrieving information from sitesHow much people use video and photo sharing sites like Flickr and YoutubeHow much they use social networking sites like Facebook and MyspaceAnd finally, how much they use RSS
- At this stage, I would like to ask for the indulgence, or should that be forgiveness, of statistical wizards for this is what you might call a rough and ready survey – but I believe it does provide some useful insights.I surveyed 90 people, well actually 100 but the last ten results came in too late to include in this presentation.I decided to break the survey into two broad groups:those who are employed on the preparation of websites.And those who just use the web either socially or for work.I also broke the web professionals into two sub-groups, the keen evangelists who I recruited at the recent WSG meeting.And web workers who are basically involved in someway or other in the preparation of websites as part of their everyday work. – These participants were recruited mainly from the offices of clients I do work for.For the web users, I surveyed 10 people from these 6 categories.I would like to stress that I did not survey anyone who I felt I knew what their likely responses would be.
- The survey contained a total of 21 questions that aimed to determine how often each participant used a particular web tool or feature, and how they used it.This is the section of the survey that dealt with blogs, and as you can see the questions seek to find out both if the participant has or uses a blog and if so, how often they use it.Basically the survey considered passive use, i.e. visiting someone's blog or looking at the photos or video of someone else. And active use, that is putting photos or videos on a photo sharing site, having your own blog or Myspace page or providing tags for some web content.After completing all the surveys, I collated the responses to each question by participants in each survey category and averaged the results. I haven’t got time to go through all of the results, so I will give a quick overview and confine myself to commenting on those which I think are most interesting.
- In order to gain a big picture view, I have averaged the results for the 11main questions in the survey for both categories of participants:Web professionls: Tools used on average by 62% of participantsWeb users: Tools used on average by 38% of participantsLET’S DRILL DOWN A BITWhen it comes to web professionals, as might be expected a greater percentage of the evangelists used all of the tools, and they used them more often, than those who were surveyed at their place of web work. 84% for web evangelists Versus 51% for web workersThere was much larger variation in the responses from the different categories of web users, with greatest use being by students where the average use of the tools was 63%. At the other end of the scale, the general public participants had an average usage of 23% and for teachers it was only slightly greater at 25%.NB: The high level of average usage by students was boosted by the fact that they had all visited a photo or video sharing site and all had social networking pages and had visited the social networking pages of others.Apart from the great difference between scores of 84% and 23%, I think there are two other things of standout interest.First, the significant difference between the evangelist and web workers, particularly when it came to taggingAnd second, the relatively low scores for teachers, media workers and to a lesser extent Australian Museum staff, since I would have thought that the nature of their work would have exposed them more to these tools and their potential benefits.
- When we look in more detail at how these tools are used, there is quite a difference between what I call active use, that is making comments, putting up material or providing tags, and passive use which is visiting or looking at the contributions of others.As might be expected, in both areas of use, the web evangelists surveyed at a WSG meeting reported much high usage. In Passive use for example, all of the evangelists said they had visited a social networking page or viewed a photo or blog of someone else, compared to about 80% for the other participants.90% of evangelists said they had used tags or tag clouds to locate something, compared to just 27% for non-web users (and I suspect even this score is a little inflated).When it came to Active use, the differences were more striking.
- As we can see here, the evangelists significantly out score the other two categories when it came to actually putting content, comments or tags onto the web. While this fact alone doesn’t surprise me, there are a few interesting things to consider:First, the disparity between the number of web evangelists and web workers who actively participate in the web. Since I would imagine all the web workers are aware of all these facilities I can only assume that the lower level of participation is due to a conscious decision and not because of a lack of knowledge.Second, the percentage of people who made comments on web pages or blogs; with 85% of evangelists saying they had done so, compared to one third of the non-web professionals surveyed. And,Third, the huge difference in the reported tagging of web content – ranging from 90% for the evangelists down to 20 and 18 percent for the web workers and people who just use the web. I suspect the 90% figure is inflated by people including the provision of tags for their own photos and video material which was not the intention of this question.There are many ways to slice and dice the figures from a survey like this, and one obvious one being gender – however I did not notice sufficient variation between the genders to warrant considering this now.But, I think it is worth making a quick mention of age differences
- Although the survey had six age categories, for the purposes of this talk I have just split the participants into two;Those 30 and under – of which there were 32 participantsAnd, those 31 and over – of which there were 58 participants.As might be expected, the younger group are more likely to have their own Facebook, Myspace or bebo type page or to be actively involved in posting photos or videos and commenting on those of others. However, these difference are not much as I would have thought.I think the relative lack of difference when it comes to commenting on web pages or blogs or subscribing to RSS is interesting. In the case of commenting, one possible reason might be that it more closely resembles the feedback mechanisms of traditional media, that the older participants are very familiar with, like letters to the editor and even talk back radio.And when it comes to RSS, established media outlets like the ABC and the Sydney Morning Herald have clearly demonstrated how useful these can be to the average punter over the last few years.
- So what, if anything, can we take from this study.First, I would like to make it clear that this survey does not set out to state that a certain tool is likely to be used by x% or the population with a y% margin of error. Rather, I am just hoping to provide some insights into the usage of these tools which some people might find useful.First, and most notably, the striking disparity between how much these tools are used by web evangelists when compared to people who see the web as just another everyday thing to be used in an everyday way. Second, this difference in usage also extends to web workers who are not evangelists – they also just see the web as something that they work with and not some all encompassing passion. To quite a large extent the behaviour of the standard web workers more closely matches that of people who just use the web rather than the evangelistsCombined, these points seem to suggest that web evangelists are not a good guide to web behaviour or how likely the general community are to adopt new ideas.Third, with reference to the “tagging” of web content, when I did the earlier survey 2 years ago 20% said they had tagged content and about 50% indicated they were likely to tag at least some web content in the future. This future must still be someway off, for in the recent survey if we exclude the web evangelists still only about 20% of participants said they had tagged content.And fourth, while many more young people are likely to have a Myspace page, age alone is not a particularly strong indicator of the likelihood of someone using these new tools. There were participants under 20 with no interest in commenting, tagging or RSS – just as there were people over 60 who did all these things regularly.In summary, when it comes to usage of the web 2.0 things I looked at, a smaller percentage of the general community than we sometimes expect use them, and those who do use them, do so much less often.
- So what should clients and developers do when considering how to maximise the usefulness of their sites. There are now many choices including:The traditional information architecture approach, with a taxonomic hierarchical system of classification and site navigationThe use of facets.And, of course the variety of social networking optionsToday I have mentioned the increasing pressure to use Google as the magic bullet, and the pitfalls of forcing potential users of your site to rely solely on web-wide search for locating information and resources. And also the dangers of being overly enticed by the allure of the new and the siren song of the latest buzz words. So, where to go?Clearly, the answer is not be found by forcing clients to choose one path. In my view, we need to adopt a mix-and-match attitude, when it comes to determining the most appropriate blend of traditional taxonomy, SEO and folksonomy. This will have to involve greater needs analysis and user testing than has often been the case in the past.I believe that social systems of classification, like tagging, do have a very important part to play in the mix for some sites. They offer users a useful associative-way of classifying content and a non-linear navigational mechanism for locating it. However, this should not be the only approach, and the challenge is how to balance up all the different pressures in order to find the best blend of techniques to meet the classification and retrieval needs of the site’s users.
- But before I consider the question of how to find the best approach, a quick quiz to wake us all up.This is a tag cloud on the Raine and Horne site, and my question is why are the names of some suburbs bigger and in a different colour.While I get this page up in Internet Explorer, write down your answer – no better still whisper it to your neighbour. We don’t want any cheating!
- In 1994, Steward Brand introduced the concept of \'Pace Layering\' in his book \'How Buildings Learn\' to describe a process of change over time. He suggested a building could be divided into \'layers\' and the speed at which the layers change varies. At the bottom layers there is the site or structural core of a building where change happens very slowly. Whereas the top layers, things like paint, fittings etc, the rate of change is the fastest.
- At the surface of the Pace Layering model, where change can happen quickly, ideas swirl around. As these filter down to the lower levels, where the rate of change is slower, they are accepted, modified or even rejected. Pace Layering theory recognises that the different rate at which change happens in each layer is likely to cause friction between layers. But, out of this friction, we learn and develop.The faster, outer layers innovate and learn while the slower core levels remember, take a long-term view and provide stability.As Brand has observed, the colour and movement of action in the outer layers tend to get the attention, but power usually resides in the slower parts.
- Steward Brand went on to apply pace-layering theory to changes in society over time in the book “The Clock of the Long Now.”Since then, Pace Layering has been applied to a wide range of situations including, study of the environment, user-centred design and, as we can see here, to information architecture by Peter Morville.
- A year ago, Russ Weakley and I got to thinking about how pace-layering might apply to the development of the web and in particular to the introduction of new technologies and the associated accessibility implications. We suggested these 6 layers for the web, ordered by the rate of change in each layer:The Internet, web and W3C are at the core and with the slowest rate of change, they provide stability and certainty. Users and developers are at the outer layers. They want to get their ideas out NOW and the rate of change is the fastest as they propose, challenge and stimulate. A lot of the Web 2.0 action happens in these outer layersThe speed at which things occur in each layer has led to some friction between the layers over the years. But this has not always been bad, in fact it probably contributed to the astronomical growth of the web.
- Ideas can ignite at one level and then spread quickly, identifying and resolving problems and gaining support along the way. In 2005 for example, Jesse James Garrett coined the term AJAX to describe a process that brought together some established technologies so developers could introduce new content to a page by re-writing sections of the DOM.It quickly became very popular: Developers liked AJAX and most users liked it, however for some assistive technology users it was inaccessible.Screen reader manufacturers quickly picked up on the potential of AJAX and became actively involved in trying to find solutions. Within about a year a year a fair amount of AJAX material could be made accessible for the users of at least one screen reader.
- As the problems thrown up by the friction resulting from the different rates of change in the outer layers are discussed and resolved, ideas flow down to the core, where standards and guidelines are developed.The core has the power to implement ideas and the stability required to make them sustainable.
- As we have seen with AJAX and, to take another example, the WAI-ARIA Roadmap, these new technologies or approaches can become incorporated into the operations of the web, and then go back out to the wider web community in the form of recognised practices, guidelines or recommendations such as WCAG 2.0.If you are interested in some of these ideas, I cover them in more detail in a rather long article called “The Evolving Web” which is on my site.
- To return to the question of how to help clients find the information classification and retrieval approach that best meets their needs and the needs of their site’s users.I think pace-layering theory offers a useful way of approaching this question, in particular because it recognises that the desire for change and the rate of change at various levels of an organisation or development process are likely to be different. And this difference is likely to result in some level of friction that will need to be managed effectively.Pace-layering, also recognises that those at the core need to allow new ideas to be explored and should be cautious when it comes to driving how things “should be done”.When considering the production of a website for an organisation, it is my view that the organisation involved, as well as the production process itself, is made up of a number of layers, each with its own influence on, or control over, the process of developing the site.
- I have represented the traditional approach to site production as a pyramid made up of 5 layers. Of course the thickness of each layer will depend on the complexity of the organisation and intention of the model.Traditionally, as suggested by this diagram, the process of developing a new site or doing a major overhaul of an existing one is often very top-down. With those at the top, senior management, the board or DG for example, having and wielding the ultimate power, and as a consequence, influence over the process.I am sure many people in this room, like me, have been involved with projects that have been compromised by arbitrary decisions of “someone up there”. Sometimes it might be the question of navigation labels where the opinion of the boss can outweigh an infinite amount of user testing or maybe it is because the CEO or the DG haven’t quite got to grips with the post Netscape 4 world. In this process, the intended users of the site are not explicitly represented and the general web community is at a level with the least influence.I now wish to suggest a pace-layer model that may apply to any medium to large organisation, be it a company or government agency, that is embarking on a major website development or re-development process.
- For this pace layering perspective, I have separated the spheres of influence over the site development process within an organisation, from the actual process of developing the site itself. I have placed this model explicitly within the context of the web, by having the web community, including evangelist and general users, at the outer layer in order to recognise the role they can have in helping to set the agenda for the development process.I have also specifically included the target audience for the site as a central layer. The reason for this is to help avoid the tendency to view web users as a homogenous group and to recognise that the needs of the users of a specific site are likely to change relatively slowly. Of course there is likely to be friction between layers when it comes to considering how best to meet the changing nature of the web. :When it comes to influence, all layers of the model can, and should, have the ability to influence the process of building or changing the organisation’s site. The ideas of the outer layers where the degree and desired rate of change is likely to be the fastest will be refined as they filter down to the coreAt the core is senior management, whose prime role should be to provide stability. Within this model, management still has the greatest power but is not well placed to force radical or rapid changes on the site development process.The main point of this model is to recognise that each layer has its own needs and perspective on the degree of change required for the site. And, that conflict between these layers in terms of the pace of change should not be avoided, but embraced in order to find the best way forward.
- This model is also useful for identifying the tensions that might exist between the layers during the processes involved in developing or redeveloping the sites. Superimposing the development processes on to the model highlights the importance of clearly identifying the different phases in the process and the desirability of adopting an iterative approach to site development with different rounds of user testing.At each phase of the development process, the relationships between one layer and another is likely to be different, in part because of the desired outcome for each particular phase and the rate of change each layer is comfortable with at that stage in order for the project to proceed.At the outset, Need Analysis incorporating the core of the organisation as well as the targeted users of the site will help refine the objectives. At this stage, it is likely the targeted users of the site will have a substantial wish list of what they would like and a desire to see their wishes fulfilled as soon as possible. It will be up to the Business analysts and marketing people to turn this wish list into achievable objectives that meet longer term strategic goals of the organisation in order to get sufficient commitment from senior management.This process should also involve the web development team, although this sometimes doesn’t occur so I just touched this layer in the diagram.
- When we move on to developing the IA and the mix of information classification and retrieval mechanisms, all options should be up for discussion including conventional site taxonomies as well as different social networking approaches such as folksonomy.At this stage, the web development team will have a key role to play and, it is likely the main line of friction will be between the developer and target audience layers. As discussed earlier, there could be pressure from some of the web developers to push the boundaries of their craft by trying the new ideas that are circulating around the web community at the time. However, these new ideas could well be challenging for the intended users of the site who might not have encountered them before. This potential friction between developers and intended users should not be viewed as a negative, but rather as an opportunity to explore alternative ways of doing things. On-going user testing of the Information Architecture and navigation mechanisms as they evolve, and discussions with the different stakeholders will increase the likelihood of resolving any differences and finding a workable solution that is acceptable to all.
- The build stage of the project will be primarily driven by the web developers and designers working to an established and tested I.A. Once again, there is the potential for friction between the layers when it comes to the desire for change and the rate of change; with the site developers perhaps wanting to go a little further and a little faster than the users of the site, or the various organisational layers nearer the core.While on the other hand, activities by web evangelists and the general web community as a whole might stimulate discussion of approaches that have not previously been considered by the developers. All options should be on the table for consideration, and ideally an iterative process of site development with a number of cycles of user testing should be adopted.As ideas are tried out and the friction between layers resolved, a common approach, with hopefully greater support from the stakeholders at the various layers, will filter down to the core and be accepted by management.
- One of the most exciting things about the web is the rapid rate of change: New ideas, new information, new technologies are appearing all the time. As web professionals, we need to embrace change while also not being blinded by the glitter of the new, for not all that shines is gold.I feel that pace-layering theory offers a useful tool for finding this balance. I have outlined one model with 6 layers, but this is just a suggestion and I am sure many of you could come up with other equally plausible models. For me, it is not the exact details of the model that are important, but the insights provided by pace-layering into how systems and entities evolve over time. Among these:The idea of looking at an organisation or process as a series of layers with differing needs for change and rates of change. An awareness that all levels of an organisation have a role to play in the process of change and renewal and the value of separating those things that change frequently from those that are more constant.Recognition that the differences in the rate of change are likely to result in friction, but rather than avoiding friction we should seek to learn from it.As we develop the Information Architecture and navigation systems for web sites we need to be flexible. With this talk, I have attempted to draw attention to the potential dangers of believing there is one, and only one, approach to classifying and retrieving information, and promote the value of exploring all options. The role of the Information Architect is to take the long term view, balancing the energy and enthusiasm of the rapidly changing web environment with the goal of making a site that will work effectively for our clients and their clients both now and into the future.
-
xiaohang favorited this 1 month ago -
a_amy28 favorited this 1 month ago
-
jamesbreeze favorited this 1 month ago
-
jamesbreeze favorited this 1 month ago
-
jamesbreeze favorited this 1 month ago
-
Added to the group Web Site Design by PatrickKennedy
-
Added to the group User Experience Design by PatrickKennedy
-
Added to the group Information Architecture by PatrickKennedy
-
kapil favorited this 1 month ago
SlideShare is now available on LinkedIn. Add it to your LinkedIn profile.
Taxonomy, Social Networks and Pace Layering
Roger Hudson discusses the roles of search, taxonomy and social networks in information classification and retrieval. Can pace-layering help us find the best approach.
748 views | 0 comments | 6 favorites | 0 downloads | embeds (Stats)
Slideshow Transcript
- Slide 1: Oz-IA - 2008 Taxonomy, Search, Social Networks and Pace Layering Roger Hudson Web Usability 21 September 2008
- Slide 2: Oz-IA - 2008 Doctor Who: “Silence in the Library”
- Slide 3: Oz-IA - 2008 Danger lurking in the shadows
- Slide 4: Oz-IA - 2008 Why in the information age do we see a decline in research skills? Why are clients handing information retrieval from their sites over to other people?
- Slide 5: Oz-IA - 2008 “Most people come to my site from Google so Information Architecture is no longer important!”
- Slide 6: Oz-IA - 2008 Developers keen to explore new heights and push the boundaries of their craft
- Slide 7: Oz-IA - 2008 Or perhaps, to out cool the cool!
- Slide 8: Oz-IA - 2008 Contradictory forces Finding the balance
- Slide 9: Oz-IA - 2008 Not about good guys and bad guys
- Slide 10: Oz-IA - 2008 Daily Internet usage: • 60% of users are using email • 49% use search engines • 39% news sites • 13% social networking sites Pew Internet and lifestyle project (August, 2008) Google has 68% of search engine share Hitwise (March, 2008)
- Slide 11: Oz-IA - 2008 What do we want from search engines? Relevance, Relevance, Relevance! Recall: Percentage of total relevant documents retrieved from all possible documents. Precision: Percentage of documents retrieved that match the searchers needs. Full recall with high precision is the aim. But who decides what is relevant?
- Slide 12: Oz-IA - 2008 This guy Newtown, did he discover gravity? Give me the facts, nothing but the f acts!
- Slide 13: Oz-IA - 2008
- Slide 14: Oz-IA - 2008 Did he?
- Slide 15: Oz-IA - 2008 “Facts are not knowledge. Information is not wisdom. And for many consumers today, more raw data does not necessarily improve their daily lives.” Professor Nancy F. Koehn, Harvard Business School “Information is not knowledge Knowledge is not wisdom Wisdom is not truth Truth is not beauty” Frank Zappa, Packard Goose (Joe’s Garage)
- Slide 16: Oz-IA - 2008 Infantilization of adults • Fast food • Fast money • Fast relationships • Shop ‘till you drop! “This is not just attention deficit disorder but compulsory attention disorder, defined by a culture in which we are dissuaded from concentration and continuity and rewarded for pursuing jump-cut lives” Benjamin Barber, “Consumed”, 2007
- Slide 17: Oz-IA - 2008 Growing reliance on web-wide search
- Slide 18: Oz-IA - 2008 Desire for instant gratification I want the answer, I need the answer, I want it NOW!
- Slide 19: Oz-IA - 2008 “The Anatomy of a Large-Scale Hypertextual Web Search Engine” Sergey Brin and Lawrence Page “The Google search engine has two important features: First, it makes use of the link structure of the Web to calculate a quality ranking for each web page. This ranking is called PageRank. Second, Google utilizes link(s) to improve search results.” http://infolab.stanford.edu/~backrub/google.html
- Slide 20: Oz-IA - 2008 “PageRank is an excellent way to prioritize the results of web keyword searches.” Sergey Brin and Lawrence Page “The “rich-get-richer” phenomenon can be particularly problematic for the ‘high-quality’ pages that were recently created.” “Impact of Search Engines on Page Popularity” Junghoo Cho and Sourashis Roy http://oak.cs.ucla.edu/~cho/papers/cho-bias.pdf
- Slide 21: Oz-IA - 2008 Most popular search queries 2001 - 2005 • Britney • Harry Potter • Brad and Angelina “The number of documents in the indices has been increasing by many orders of magnitude, but the user's ability to look at documents has not. People are still only willing to look at the first few tens of results.” Sergey Brin and Lawrence Page
- Slide 22: Oz-IA - 2008 “The ACCC is alleging that Google, by failing to adequately distinguish sponsored links from \"organic\" search results, has engaged and continues to engage in misleading and deceptive conduct.“ Australian Competition & Consumer Commission,July 2007
- Slide 23: Oz-IA - 2008 “I predict that liberation from search engines will be one of the biggest strategic issues for websites in the coming years. The question is: How can websites devote more of their budgets to keeping customers, rather than simply advertising for new visitors?” Jakob Nielsen’s Alertbox, 2006
- Slide 24: Oz-IA - 2008 Bring on the folk! “The advantage of folksonomies isn’t that they’re better than controlled vocabularies, it’s that they’re better than nothing.” Clay Shirkey, 2005
- Slide 25: Oz-IA - 2008 Social software “A folksonomy is a set of uncontrolled tags provided by individuals for their own retrieval purposes of that object and these tags are shared publicly.” Thomas Vander Wal
- Slide 26: Oz-IA - 2008 Folksonomy and Tagging • ‘Folksonomy’ is an open-ended labelling system that allows users to categorise online content. • Users provide descriptive keywords or ‘tags’, which use familiar, shared vocabularies. Folksonomy is the sharing of tags provided by different users. Assumption: If enough people tag an object, interesting and useful patterns will emerge.
- Slide 27: Oz-IA - 2008 Looking for answers: 2006 survey Survey participants: • 10 media workers (radio reporters and producers) • 10 library workers (librarians, archivists and researchers) • 10 web workers (producers, designers and developers) • 10 museum workers (scientists and program managers) Key questions: • Are they aware of tags and social booking marking? • What sort of tags might they use? • How likely are they to tag in the future?
- Slide 28: Oz-IA - 2008 Tag these photos 49 different tags, 29 unique 67 different tags, 47 unique
- Slide 29: Oz-IA - 2008 At the end of the survey each participant was asked: “If in the future you could provide tags for web content (pages, images) that might be helpful to you and other users, how often would you do this?” Never Infrequently Sometimes Often Always 4 15 10 6 5 Comments include: • I just want to get the information and get out. • I might if it helps other people. • Don’t have the time. • What’s in it for me?
- Slide 30: Oz-IA - 2008 Pushing the boundaries
- Slide 31: Oz-IA - 2008 “The Brooklyn Museum believes in community and in the importance of the visitor experience. In this area you'll find a number of ways to connect with us: blogs, photo and video submissions, podcasts, and more. We look forward to hearing from you.” www.brooklynmuseum.org/community/
- Slide 32: Oz-IA - 2008 ....
- Slide 33: Oz-IA - 2008 Web Tools Survey: 2008 Aims: Obtain an insight into the usage of some of the more recent web tools. See if there is any difference in the usage by web workers and general users of the web.
- Slide 34: Oz-IA - 2008 Web Tools Survey: 2008 Issues considered: • Providing comments on web pages • Blogs • Tagging content and using tags to retrieve content • Use of video and/or photo sharing sites • Use of social networking sites • Use of RSS
- Slide 35: Oz-IA - 2008 Web Tools Survey: 2008 Participants (n=90) Web professionals (n=30) • Evangelists (WSG meeting attendees) • Workers (employed on website production) Web users (n=60) • Australian Museum staff (scientists and project officers) • Amnesty International Australia (volunteers and staff) • Teachers (high school and primary school) • Tertiary students • Media workers (ABC journalists and broadcasters) • General public
- Slide 36: Oz-IA - 2008 Web Tools Survey: 2008 Overview of questions 21 questions to determine how a particular web tool or feature is used. For example;
- Slide 37: Oz-IA - 2008 Web Tools Survey: 2008 Overall results Average usage of all tools by survey categories: Web professionals: Used by 62% of participants • Web evangelists: 84% • Web workers: 51% Web users: Used by 38% of participants • Tertiary students: 63% • Media workers: 43% • Museum staff: 35% • Teachers: 25% • General public: 23%
- Slide 38: Oz-IA - 2008 Web Tools Survey: 2008 Passive use (% of participants) WSG Web Non-web Evangelists Workers people Visited photo/video 100 85 85 sharing site (eg Flickr) Visited someone's page 100 60 72 on social network site (eg Myspace) Used tags or tag cloud 90 65 27 to find web resource
- Slide 39: Oz-IA - 2008 Web Tools Survey: 2008 Active use (% of participants) WSG Web Non-web Evangelists Workers people Made comment on web 85 55 34 page or blog Posted photo/video 70 45 22 Commented on 60 50 32 photo/video (eg Flickr) Own social network 100 60 55 page (eg Myspace) Tagged web content 90 20 18
- Slide 40: Oz-IA - 2008 Web Tools Survey: 2008 Use by age difference (% of participants) 30 yrs 31 yrs or less or more (n=32) (n=58) Made comment on web page or blog 56 40 Posted photo/video 44 28 Commented on photo/video (eg Flickr) 62 29 Own social network page (eg Myspace) 97 40 Tagged web content 44 17 Subscribe to RSS 37 34
- Slide 41: Oz-IA - 2008 Can we learn anything from the 2008 survey?
- Slide 42: Oz-IA - 2008 Where to go?
- Slide 43: Oz-IA - 2008 Time for a quick tagging quiz What determines the size/colour of the suburbs?
- Slide 44: Oz-IA - 2008 Pace-Layering Stuart Brand introduced “Pace Layering” in his book “How Buildings Learn: What Happens After They’re Built” Fast rate of change Stuff Space plan Services Skin Structure Site Slow rate of change
- Slide 45: Oz-IA - 2008 Pace layering - key elements • Changes that happen at one layer filter down • Differing rates of change cause friction between layers • The different rates of change are beneficial \"The fast parts learn, propose, and absorb shocks; the slow parts remember, integrate, and constrain. The fast parts get all the attention. The slow parts have all the power.” Steward Brand, The Long Now Foundation
- Slide 46: Oz-IA - 2008 Pace-Layering in Information architecture society layers Fast Fast Fashion & art Content, services, interface Commerce Adaptive finding tools Infrastructure Controlled vocabulary Governance Enabling technologies Culture Embedded navigation Nature Faceted classification Slow Slow “Clock of the Long “Speed of Information Now”, Stewart Brand Architecture”, Peter Morville
- Slide 47: Oz-IA - 2008 Pace layering and the web
- Slide 48: Oz-IA - 2008 Pace layering and the web
- Slide 49: Oz-IA - 2008 Pace layering and the web
- Slide 50: Oz-IA - 2008 Pace layering and the web
- Slide 51: Oz-IA - 2008 Can pace-layering help clients and developers determine the best approach?
- Slide 52: Oz-IA - 2008 Site Production: Traditional Levels of Decisions & Influence Most influence and power Least influence and power
- Slide 53: Oz-IA - 2008 Spheres of influence Pace-Laying view
- Slide 54: Oz-IA - 2008 Processes Pace-Laying view
- Slide 55: Oz-IA - 2008 Processes Pace-Laying view
- Slide 56: Oz-IA - 2008 Processes Pace-Laying view
- Slide 57: Oz-IA - 2008 Conclusion Change is constant
- Slide 58: Oz-IA - 2008 Thankyou Roger Hudson Web Usability 0405 320 014 rhudson@usability.com.au
