A Case For Knowledge Base Driven Information Mining – A WebSummarizer Use Case

We use terms like “knowledge economy” (Peter Drucker) , “knowledge as the fuel for innovation” (Nonaka) to emphasize the value we put on knowledge.

Putting knowledge to work requires tools enabling the gathering and structuring relevant information as factors driving innovation, creativity, and  invention.  In practice tools and techniques are tangible instruments of applying knowledge to solving real life challenges.

This is especially important in modern information society “where the creation, distribution, use, integration and manipulation of information is a significant economic, political, and cultural activity”. (see here)

Both concepts - knowledge economy and information society - are closely related by its emphasis on increasing technological capacity to store,  analyze and compute information as bases for data mining and discovery.

In our daily practice we can see the that a direct impact of knowledge economy is measurable in the explosion of scientific and medical discoveries, technological  patents and rapid research progress.

Information overload and information wealth

On a practical human scale we are facing information overload – a condition described as “the difficulty a person can have understanding an issue and making decisions that can be caused by the presence of too much information.” (see here)

Clay Shirky operationally defined information overload rather as a case of filter failure in his famous talk at  Web 2.0 Expo NY:  "It's Not InformationOverload. It's Filter Failure"

In reality we should rejoice and take advantage of the fact that we have unprecedented in the human history access to information wealth. And to cope better with information wealth we need automated  smart tools allowing us to take advantage of this unparalleled access to information.

Let’s take a practical case from our daily work. Research tells us that we spend almost  one third of our time searching for relevant information.  Of course finding relevant data is critical to our productivity but  spending one third of our productive time on searches leaves not much time on the actual creative usage of our findings.

On one hand we have the challenges with large data sources that includes capture, storage, and search of relevant information and then we need effective tools for sharing, analysis, and visualization. On the other hand all of us can benefit from knowledge mining tools allowing us to concentrate on applying knowledge instead of using our precious time on inefficient searches. 

Knowledge Base – practical tool set for discovery

In short knowledge base allows us to see how different pieces of information are inter-connected! This is the principle of how the Web works: it is all about linking web pages.

However, what is even more useful is to link the topics that are discussed on those web pages with related topics on other pages. This allows are to purely look at the relationships as a family of connections without the distraction of seeing the the unrelated information.  This allows us to focus on the topic in context of its relationship, correlation, influence, meaning, to other topics and contexts. Such information filters help with understanding and conveniently provide facts and insights to form our judgment and decisions.

Here is an example of Visual Knowledge Map. The topic is "Learning". The sources for generating the Visual Knowledge Map are numerous articles in Wikipedia that address the subject of learning in rich context. The Visual Knowledge Map was generated by WebSummarizer.

The Visual Knowledge Map is interactive. You can click on the + and - signs to zoom in and zoom out.

 

Visual Knowledge Map - Topic: Learning - Generated by WebSummarizer from Wikipedia knowledge base

Visual Knowledge Map - Topic: Learning - Generated by WebSummarizer from Wikipedia knowledge base

 --------------------------------------------------------------------------------------------------- 

About WebSummarizer:

WebSummarizer is a powerful text mining and visualization application.  It offers rapid summarization of web pages and documents, creation of personal and corporate knowledge bases. 

The summarization results are presented as: 

•   Visual Summaries

•   Visual Knowledge Maps

•   Tree Views (structured text)

•   Keywords Cloud, and

•   Visual Summaries and Visual Knowledge Maps can be exported to HTML, word editors and mind mapping applications. 

You can summarize text in English, French, German and Spanish.

         VISUALIZE and SUMMARIZE web pages and documents with WebSummarizer.

BlogSummarizer, WebSummarizer and WikiSummarizer are products of Context Discovery Inc.

The Power of Visual Summary

As Internet users we spend a significant amount of time examining search engine results.

For each search results we must click through many links to briefly evaluate each for possible relevance to a particular information we need. Improving the efficiency of this tedious process directly benefits the users.

The standard practice is to provide 1 sentence textual summary snippet by the search engine.  Specifically, Google shows up to 156 characters (including spaces) of a page’s meta description tag.

Here is a typical example of a Google summarization snippet:

Google summarization snippet

We believe that this practice can be significantly enhanced by providing the ability to automatically generate a more detailed visual and textual summaries of web pages.

Visual summaries greatly increase the efficiency by which readers evaluate and use search engine results. For example, keywords in visual summaries allow users to instantly see the most significant content at once without searches.

This visual presentation of the most pertinent information is a tremendous time saver. 

Let's look at an example of a Visual Summary for the "coffee"search term generated by WebSummarizer:

Visual Summary example

 
Advantages of Visual Summary presentation:

• Visual summaries conveniently aggregate the most important information around the key ideas. This makes it very easy to examine summaries by keywords.

•  In addition there are 2 options: Short Summary and Full Summary. Both summaries provides the most relevant extracts for quick orientation.
  

•  Visual Summary format enforces readability and with the highlighted keywords enhances quick comprehension of the textual summary.

• Visual presentation enhances ability to remember. It pairs the visual presentation with structured outline to make it easy to reuse the information when creating new documents.

• Providing the ability to expand and collapse the Visual Summary branches gives users great freedom to zoom in only on the points of interest.

The Visual Summary presentation makes it easy to understand and remember the key information.

Here is the dynamic Visual Summary of the coffee Wikipedia article:

                               VISUAL  SUMMARY                           

 

  
To navigate the Visual Summary:  
* click on any + or - next to a keyword in the map  
* or click on the Expand All or Collapse All buttons. ----------------------------------------------------------------   

 A bit of science

• Visual information presentation speeds up many tasks tremendously. We can get the gist of an image in 110 ms or less! (Coltheart, 1999)
• In that same 110 ms, we can read on average less than 1 word or skim two words. The average reader of English reads about 4.2 words per second and can skim or scan at roughly 17 words per second. (Chapman, 1993) 

References:

• Coltheart, V. (Ed.). (1999). Fleeting memories: Cognition of brief visual stimuli. Cambridge, MA: MIT Press.).

• Chapman, A. (Ed.). (1993). Making sense: Teaching critical reading across the curriculum. New York: The College Board.).

-------------------------------------------------------------------------------------------  
About WebSummarizer  
 

WebSummarizer provides tools to VISUALIZE and SUMMARIZE web pages and documents.

The results are presented as interactive Visual Summaries,  Tree Views and a Keyword Clouds and can be downloaded easily published in blogs and websites

The Visual Summary can be navigated in any browser on Windows, Mac, Linux, iPhone, iPad and Android devices. The Keyword Cloud is linked with Wikipedia Knowledge Base. When you click on the keyword in the cloud you will be presented with an instant Visual Summary.

You can summarize text in English, French, German and Spanish.

The keywords and summaries are easily exported to other applications such as word editors, browsers, mind mapping applications like Mindjet MindManager, MindGenius, XMind, and any other mind mapping application.

FREE Trial here: WebSummarizer

Keywords and summaries are a quick test of informativeness

Illustation by Stephen Doyle

Writing in the Internet age is a particular form of  communication art where informativeness - the significance of provided information - must be delivered directly and without fluff. 

Linguistic studies show that there is a direct relation between language of informativeness and success or failure in keeping the reader’s attention.

Any indulgence in verbosity dilutes the message and tires the reader.

Reading is type of filtering and sorting exercise.

While we are reading, often scanning and skimming through the text, we are looking for words that catch our attention. These keywords, and often key phrases, serve as a highly condensed summary of the text.
Key phrases are often highlighted within the body of the document to facilitate users’ fast browsing and reading.

In the academic practice for publishing papers key phrases are usually manually assigned by authors, especially for journals, conference articles or back-of-the book indexes. However, the vast majority of documents such as news, magazine articles, web pages do not provide key phrases because the manual effort of assigning key phrases is too expensive.

It stands to reason that applying intelligent automated tools for summarizing is practical and helpful in improving  communication and speed up understanding.

One of such tools is WebSummarizer. Here is a sample of WebSummarizer’s Visual Summary based in this post:


 

To navigate the Visual Summary:
 * click on any + or - next to a keyword in the map
 * or click on the Expand All or Collapse All buttons.   
 
As we can see in this example, when we condense the text to a set of keywords and summaries this will provide an important measure of informativeness. It tells the reader what are the main topics and in what context these topics are discussed. This informativeness index is like a cognitive tool helping the readers to quickly gauge their interest in the text.

Moreover, readers can easily examine each keyword and the related summaries to quickly glance the details. The visualization of information  is one of the key advantages of Visual Summary presentation.

By visual examination of the key phrases and summaries it is much easier to make decision whether to read the piece in depth or just skip it.

Practice shows that  it is useful to automatically extract key phrases from web pages to accompany the main content of the document. And with automated tools such as WebSummarizer providing summaries is very easy.

Key phrases are all about being informative! 

----------------------------------------------------------------
About WebSummarizer

WebSummarizer provides tools to VISUALIZE and SUMMARIZE web pages and documents.

The results are presented as interactive Visual Summaries,  Tree Views and a Keyword Clouds and can be downloaded easily published in blogs and websites

The Visual Summary can be navigated in any browser on Windows, Mac, Linux, iPhone, iPad and Android devices. The Keyword Cloud is linked with Wikipedia Knowledge Base. When you click on the keyword in the cloud you will be presented with an instant Visual Summary.

You can summarize text in English, French, German and Spanish.

The keywords and summaries are easily exported to other applications such as word editors, browsers, mind mapping applications like Mindjet MindManager, MindGenius, XMind, and any other mind mapping application.

FREE Trial here: WebSummarizer