Turning information overload into useful knowledge resources

Jan Wyllie, jan@trendmon.demon.co.uk wrote this article in 1997.

Email and World Wide Web have great attractions - but the sheer volume of information can become intimidating. Jan Wyllie offers some solutions.


Once people become connected through email to (multiple) groups of common interest -- yes, tele-communities of interest -- they find it irresistible. They want to converse with everybody. A few people become addicted to tele-conversations which begin to eat up more and more of their time to the detriment of the 'real life' (known as 'RL' in cyberspeak). People can become quickly overwhelmed. Then, one of two things can happen.

They either suffer from having to do an ever more onerous chore, or they stop replying to anyone but an inner core of associates.

The same logic applies to the availability of information on the World Wide Web: it either makes people feel overwhelmed, guilty and inadequate for not being able even to begin to keep up, or they try to forget that this enormous and ever growing global repository of knowledge exists.

The key to changing information from being an unwanted overload to being a valuable resource is good information management practice. Most expert practitioners believe that classification and ordering add value to an information base. Simply by clustering information about similar topics, it is possible to see and reflect on threads of interpersonal communication from a higher perspective making it possible when required to rise above the confusion.

Fortunately over the past 15 years, computer software, such as text databases, computer conferencing, mail filters and search agents (see below), have appeared which provide information managers with very powerful tools.

As for the time problem, the good news is that like in many professional activities, lay people can learn to do the bulk of the work, very quickly. Most good practice is based on common sense and uses natural human skills, such as the ability to categorise and distinguish levels of generality.

Software tools for information management

Text retrieval / text database

A text database will index every word in every one of thousands-to-millions of documents making each one instantly retrievable simply by typing in combinations of words it contains. Suddenly information which was safely out of reach in places like libraries is available directly from the desktop.


The drawback is that searches can either swamp users with much too much information, or completely miss what the user is looking for. To avoid this use free text retrieval combined with traditional classification schemes and controlled description vocabularies.

Computer conferencing


Computer conferencing enables geographically dispersed groups of people to carry on written conversations 'threaded' by topic. People can contribute to these electronic conversations at any time of their convenience ('asynchronous' unlike 'synchronous' videoconferencing, for example, where everyone must be there at the same time) either by starting topics of their own, or replying to an already existing conversational thread.
The benefit is that people can communicate with each other in their own time in a consistent and focused way. Results should be more useful information exchange, more informed decisions and better aligned actions.

The danger is that the signal to noise ratio is very low as chat, confusion and repetition take over from focused collaborative thinking and discussion. In order to avoid this problem groups need to agree some structural elements to their conversations. e.g. major topics and sub-topics. Moderators need to be appointed to stimulate conversation through effective questioning and to make sure people stick to agreed topics. Topics must be periodically summarised in order to give existing members overviews of what they said and to update new members.

Filters / search agents

Filters select and screen out items from email and text databases such as the World Wide Web by comparing document text with user-defined word combinations.

They can reduce the number of items to look at - however filters can screen out items for which the user never thought to ask. Agents can bring simply bring in more unwanted material ironically causing greater expense in both time and money.

These tools should be used, if possible, in conjunction with classified information bases, as well as being amended to take into account new vocabularies and information types -- a process which takes a lot of expert human work. Also, steps must be taken so that serendipity is not screened out of the information acquisition process.

Trend Monitor International Ltd., Portsmouth


Home | Communities | Networks | Partnerships | Toolkit | | Join in

www.partnerships.org.uk/articles/over1.html