Areas, Ask and Google Minds

The idea of creating and storing "know-how", and using users' questions and answers is both old and obvious.

Arguably, the most prominent example is The system's been on the market for 15 years and is heavily promoted by partners and investors like Oracle. Any time you install JAVA there is a default option to install an "Ask" toolbar into your browser. Still, you've probably never used it and have never asked anything.

The "Ask" system isn't used because it's not natural. And technologies usually fail if they require unnatural behaviour.

Now, you're probably wondering, how come questions and answers are unnatural? A fair question. We need to jump back 10 years, to the Medieval age, when the internet wasn't so widespread, to answer that. What did you do ten years ago, when you wanted to find out about something, like for instance, how to fix a squeaking gearbox? You'd call your cousin who was a car maniac, or you'd go to a technical library.

This is different to the "Ask" system. Questions and answers are, usually, highly interactive. Quite often it's a discussion. Consider a company meeting or a help-desk phone call. The "Ask" system is not.

Questions and answers are also, mostly private. Consider asking a question in public, say at a conference, compared to asking a friend in the pub. The former requires far more courage and effort. You think about how to formulate your question, you consider how you will look and sound. The latter requires almost no effort or preparation, words flow freely.

Questions and answers are often a solitary search. Think about going to a library or using Google.

Nowadays, instead of a technical library we have the internet and instead of a library catalogue, we have Google search. What have we got instead of our cousins? Clearly it's not The most similar behaviour is when we use Google to find a relevant blog or chat room, both of which parallel "going to the pub, sitting down, listening and eventually asking", or "going to public lectures".

So if, asking questions and answering are interactive activities, you'll hesitate to ask a question using some anonymous, black box. Especially if your question is to be made public, running the risk that you'll look like an idiot -or if you are professional - worse, look like an amateur. Who wants to risk a quote like "He asks such basic questions about SQL? He should know after ten years in software business ..."? That's why and similar systems are doomed from the outset and will remain marginal services no matter how aggressively they're promoted. 

So Google is the answer to everything, is that what I'm saying? Errm, actually, no, no I'm not.

Google and the Google Mind - in a decade the word has gone from mathematical obscurity to being one of the most used in the English language. We probably all have a search box in our browser's toolbar. But even though we all use it everyday, does it really answer our needs? Let's jump back a decade again.

Traditionally people used to carry in their minds, extended social maps. It meant that within their locality/ world they knew who knew what and who knew whom. Nowadays we don't consult these internal social maps. Instead of a cousin we consult Google. Where's the harm in that? Well, psychologists have compared the size of these contemporary Google mind maps to what we carried around in our skulls ten years ago and found measurable and significant losses.

Google mind is our mind without social-knowledge maps.

So where does that put us? We don't use and similar systems because they're unnatural. And Google, although it appears to release us actually puts limitations on us.

Is there an alternative, a third way?

My gut feeling is that the answer is yes.

If technology destroyed the social maps in our brains, it could be used to put them back and even enhance them. Areas, which are linked to people and their qualifications, could be used to create computerised, social-knowledge maps without needing forced user participation. Such a map could be created and updated automatically as a byproduct of other user actions in a knowledge-base system - in our case, Ency.

Similar to Google and the Internet which have become increasingly computerised extensions of memories and recollections, an "ask" map could be a computerised extension of our diminished social-knowledge maps.

What are the benefits to this third way? Why would a company consider investing in it? Firstly, employees would be able to share knowledge more effectively and thus save time and increase the quality of their outputs. Secondly, management would know who is an expert in which area and how much value would be lost if they were to lose an employee.



How would it work?


It's actually quite simple. ESKM or the Ency Social-Knowledge Maps would actually be low-hanging fruits and could be drawn easily by using already-existing data.

To start drawing our map, or ESKM, in our system, let's start with the assumption that a person is an expert if he/she actively participates in some areas which are linked (within our system) to specific projects, or terms that indicate an expertise or specific knowledge. That's a bit eye-watering but I hope what I mean will become clearer in a minute.

Our second assumption is that "expertise" can be measured and given a value. Our individual's "expertise" weight would be calculated based on the number of links (mentioned above) he/she has. The more links between a person and a term that requires expertise, the more of an expert in said term, that person is.

What do I mean by that? Let's say, we have a term, X, that requires some expertise to be used properly. Within our system there's a project Y, that uses X alot. If there is a individual who contributes to the pages of Y, the links and comments he makes to Y's pages, will also contribute to qualifying him as an expert in "X" to some degree - because logically, to know about Y he needs to be an expert in X. If he also contributes to a different area where expertise in "X" is required then his probability of being an expert in "X" will have been increased even further.

For example you have an area that is in fact a project, say "Customer Churn Report". This area is linked to the "Customer DW Mart" term and this is also linked to the "Teratadata" term. In this case, a person who creates a page in this CCR area would get an "expertise" weighting of 100 percent probability connected to it (sure he is; he wrote the article about it), an "expertise" weighting of 50 percent to "Customer DW Mart" (there's a good chance that he'll also know something about this closely-related topic) and 25 percent to "Teradata" (he may know something about it or know somebody who does).

These weighted "expertise" values generated from their links would be summed up for each person. The list of relevant persons would then be displayed, for example as an ordered list linked to the Term.

From the user perspective, it would work something like this: You have some problem and you need an answer or need to know something about it. Search Ency, find a term related to your problem or question, click on the "Experts" button and you'll get a list of people who may know something more than you do about this damned thing.

Another way of looking at it is, think of an ESKM as a Linkedin network that changes according to the topic (problem, area, expertise) you select.

Granted, this approach has some potential problems. We can't know for sure what exact knowledge our "expert" has before we ask them to help us. And it might actually discourage people from contributing to pages and projects in Ency if they become worried that they'll become "experts" and be buried under an avalanche of future questions from "non-experts" seeking solutions. But those kind of issues are implementation and management problems and aren't in the scope of this blog.

We set out to ask if there weren't alternative ways to asking and answering questions in today's "Google" age and I hope I've succeeded in outlining one possibility. If you'd like to find out more about Ency and ask a question of our own resident "experts" you won't need a social-knowledge map to get there.