The Future of Results Pages? Options

[ILoveJackDaniels]

Interesting Article about search results and the danger we have of being locked into an imperfect system of displaying them (because it may become too difficult to change). The Qwerty keyboard is a good example of how this can happen - it is now used only because it is widely used, despite the existence of superior alternatives like Dvorak, and despite the reasons for the Qwerty being used are now irrelevant.

Does anyone have any good resources on Data Visualisation? There are a few engines offering better display of search results, but I've not seen any examples of what other techniques there might be that haven't been tried yet ...

[sebastienbillard]

You may want to check http://kartoo.com/

I personnaly dislike this visual stuff, but some people do appreciate it, see for yourself

[invader]

You know there's a Firefox extension called "GooglePreview"
that ads on a thumbnail snapshot of each result in the SERP's.
Not bad.

[bragadocchio]

Great topic.

Does anyone have any good resources on Data Visualisation? There are a few engines offering better display of search results, but I've not seen any examples of what other techniques there might be that haven't been tried yet ...

I started searching for some papers onthe subject, and there are a few out there. Here are a few that cover different concepts and show a little bit of the range of potential answers to your question.

One thing that we have to keep in mind is that there are people connecting to the web using phones and other handheld devices. What is the best search interface for them?

Sorting out Searching on Small Screen Devices (pdf)

This OK/Cancel article doesn't look at new visualizations, but it's a nice comparison of ones we might be familiar with:

The Perfect Search Engine Interface

This one isn't terribly new, but it doesn give a little insight into how clustering can work in different ways:

Grouper: A Dynamic Clustering Interface to Web Search Results (pdf)

Since I mentioned some of the potential problems of visualizing search results on small screens, I thought it might also be interesting to consider the best way to display search results for people who can't see them:

Accessibility and Usability of Search Engine Interfaces: Preliminary Testing (pdf)

We also make some assumptions about how people search, and those assumptions may not always be right. I often don't search for topics by using related keywords, but rather by what I remember about where the information may be located. I may not be alone:

The Perfect Search Engine Is Not Enough: A Study of Orienteering Behavior in Directed Search (pdf)

One of the interesting points that it makes is the importance of showing the context of results as they are displayed by a search engine.

I also took a look at a search engine mentioned in a comment on the OK/Cancel site that was pretty interesting. The Language computer didn't do too well for questions like "How's it going? and "Who's your daddy," but it gave me some great answers for "Who won the U.S. election for president in 2000?"

http://www.languagecomputer.com/scripts/question.cgi

[Grumpus]

The notion that the way we read search results may lock us into this way rather than a new way is complete rubbish. Comparing it to a Qwerty/Dvorak or (as the article did) Imperial/Metric Systems analogy just doesn't work either.

Both typing and measuring are learned skills. If you suddenly switched to the Dvorak system and eliminated the Qwerty boards around the world, productivity would suddenly drop to almost nothing in almost every industry. Only after months and months of constant training and practice could productivity be returned. The cost of that training would be immense and the return would be (I dunno, I don't have numbers on this, so I'll make some up) maybe, at best, a 10% increase - probably a lot less. Just doesn't make sense.

The same is true with the metric system. Even though base 10 is much easier to work with, you still need to learn all the names and figure out that if a "mil" is a thousand, then what the heck is a million? You also need to have some sort of visual comparison. I know how much a liter is because I buy my soda pop in liter or 2 liter bottles. I can look at a large tank and estimate that it's about 400 gallons, but my comprehension of a liter won't help me figure out how many liters that same tank is. So, I've got to figure it out.

Both of these examples are rather equivalent to learning a second language. It's not something that you can just say, "I'm going to do," and then wake up the next morning and do it.

Reading search results isn't a learned skill. Creating a usable search is a skill, but reading the results is nothing but obvious. You could change tomorrow and I might yearn for my old familiar system, but if the new system is effective and intuitive, it's not going to affect my productivity in the least.

So, the problem is in the way they are trying to give us data visualizations. I've never seen a method that I particularly like. They aren't particularly intuitive and they don't save me any steps or time. With the linear search results, I roll the dice and look. I'm presented with a list sorted by the odds that that document contains what I'm looking for. I can scan that list and if it looks like nothing is going to meet my needs, I rethink my search and roll the dice again. With a visualization engine, I roll the dice and then have to make another decision to refine those results in some way before I can get to my answer. Even if you manage to come up with a good way of visualizing results, you still always need to make a second choice to move in and get to your answer. It's not particualrly productive and, therefore, may not always be better than the current way of showing results.

There's also nothing saying that you can't offer both. Here's the old way. No? Doesn't look good? Okay, now pick from one of these and we'll see if we can nail this down for you. So far, all of the engines have been exclusively one or the other.

So, this is really just a case of not giving me a viable alternative. When I lived in Newport, RI all of us locals rode around on Mopeds all summer because the streets were so crowded with tourists, it wasn't uncommon for it to take us 20 minutes to drive two miles. With Mopeds, we could make that same trip in under 5 minutes. But, this didn't help when you were planning on leaving the island. True, my moped was the most efficient way to get to the bridge, but once I got there, that moped lost all of its usefulness because the traffic thinned and now I was moving at a quarter the speed of the people in cars - and I am only 10% or so of the distance along in my journey.

In essence, the danger we have of getting locked into the current means of reading search results is akin to the act of telling me that I have but one choice - drive my car all the time, or drive my moped all the time, but I don't have the option of choosing the best (easiest, most efficient, most productive) one for the particular task I have at hand.

G.

[cre8pc]

See also:

Changing the Way People Search for Information, Through Algorithms and User Interfaces

Senior Microsoft Corp. Researcher Susan Dumais predicts that in 10 years, we will look back on today'ss search interfaces and recognize them as a simple and limited way to interact with information. After all, she explains, a 5-inch-long rectangle with a long list of text results beneath it doesn't do much to help people make sense of the billions upon billions of unorganized bits of data in the world.