Knowledge Representation: Utilizing Mind Maps as a Structure for Mining the Semantic Web

    In the past 2 and half years I followed the Online Masters Programme of University of Liverpool, it was intriguing, fun, time consuming and quite an effort as energy, money and personal life, and I hope it will pay back in time, the sooner the better. The modules were quite entertaining, I learn lot of new stuff and two years passed fast and slower than expected, then the dissertation came and things got pretty tough as I wanted to make it useful for me, to learn something meaningful on which I can built in the future and not something that will rot in a corner of the brain. I was not sure what to choose and frankly not even what I supposed to do.

    During the last modules I had the chance to read some material on Tony Buzan’s Mind Maps and it looked intriguing, I wish I had have read that stuff long time ago, but in the end better later than never. Why I found Mind Maps intriguing? First because they allow taking notes in a radiant fashion rather than using the old fashioned linear approach, by starting with a single idea (also referred as concept, subject, question) and built around it a whole Map using associations. In Mind Maps Key-Words are used to encapsulate a variety of meaning in smallest possible units, this step allowing some information filtering and processing, “obligating” the brain to actually integrate the new information in existing knowledge and represent already existing knowledge, identifying missing links, triggering other questions, etc. Thus on a piece of paper or in a electronic document, somebody can represent how concepts in a read material link to each other, making the subject clearer and I think easier to memorize and recall. Mind Maps can be also used to give life to own mental representations, as we all have created, voluntarily or involuntarily and map of the world we live in. Secondly, Mind Maps use symbols and graphical images, visual rhythms and patterns, colour and spatial awareness (dimension and gestalt), allowing people to take advantage of a broader set of cortical skills.

    Given these characteristics, Mind Maps seems to be perfect tools for Knowledge Representation in particular and Knowledge Management in general. During the Web Applications module I tangentially learned about XTM (eXtensible Topic Maps) and ontologies for Knowledge Representation, though ontologies call for experts and come with many issues, while XTM is a standard for Knowledge Interchange and targets internal representation in computers. On the other side digital Mind Maps are more flexible than ontologies, target a broader range of users, have the potential of harnessing the Collective Intelligence, one of Web 2.0’s competences, by allowing users to map their knowledge or the knowledge existing on the Web in documents. This is how appeared the title of my Dissertation paper, “Utilizing Mind Maps as a Structure for Mining the Semantic Web”.

    While diving in the subject, I found out that Mind Maps are just one of the Knowledge Maps used for various tasks, a search trough the literature revealing about 50 terms used to designate various types of Maps: argument maps, brace maps , bridge maps, bubble maps, causal maps, circle maps, cluster maps, cluster vee diagrams, clustering, cognitive maps, concept circle diagrams, concept maps, conceptual graphs, congregate maps, diagnostic maps, double bubble maps, dynamic cognitive maps, ecological maps, extended fuzzy cognitive maps, flow maps, frames, fuzzy cognitive maps, fuzzy relational maps, group maps, historical maps, idea maps, knowledge maps, mental maps, mind maps, multi-flow maps, neural cognitive maps, neutrosophic cognitive maps, node-link mappings, ontology, oval maps, probability fuzzy cognitive maps, rule-based fuzzy cognitive maps, semantic maps, semantic nets, semantic networks, shared maps, social maps, social mess maps, spider maps, strategy maps, taxonomy, text graphs, thinking maps, tree maps and virtual maps. Actually, the list might be much bigger, I expect I left out by mistake several terms, while on others I haven’t came across them until now.

    From several considerations, I preferred to treat the subject from the perspective of Knowledge Maps, so maybe a better title for my paper would have been “Utilizing Knowledge Maps as a Structure for Mining the Semantic Web”. As I found out later this cost me a huge amount of time and effort, I longed for more I could chew in the dedicated amount of time for a Dissertation paper, not having the time to bring the paper to the desired final form, letting out some research material and ideas. Anyway, now it’s over, good or bad the paper is finished and waiting for the final results. With this blog I’m hoping to bring into light some of the ideas I couldn’t put in the paper, help me do to further research into the subject and hopefully get also some feedback.

    I’m not sure yet whether I can put the paper in the public domain, therefore here is paper’s Table of Contents, with the mention that some of the topics (e.g. Fuzzy Cognitive Map) have only an informative character.

1. The Web

1.1 Introduction

1.2 Web 2.0

1.3 The Semantic Web

1.4 Semantic Web Problems

1.5 Beyond the Semantic Web

1.5.1 The Noosphere

1.5.2 Cognitive Machines

2. Philosophical Grounds

2.1 Introduction

2.2 From Meaning to Concept

2.3 Syntax, Semantics and Pragmatics

2.4 From data to wisdom

2.5 Types of Knowledge

2.6 Connectivism

2.6.1 Introduction

2.6.2 Chaos

2.6.3 Network

2.6.4 Complexity

2.6.5 Self-organization

2.7 Intelligence and Collective Intelligence

2.7.1 Intelligence

2.7.2 Collective Intelligence

2.7.3 Collective Web Intelligence

2.7.4 Web Technologies and Collective Intelligence

2.7.5 Offline Collective Intelligence

2.7.6. Collective Intelligence Forms

3. Knowledge Management

3.1 Introduction

3.2 Knowledge Representation

3.2.1 Introduction

3.2.2 Sub-conceptual level

3.2.3 Symbolic level

3.2.3.1 Generalities

3.2.3.2 The Frame Problem

3.2.3.4 The Symbol Grounding Problem

3.2.4 Conceptual level

3.2.5 Associationist level

3.2.6 Semantic level

3.3 From Mental Models to Knowledge Representation Structures

3.4 Historical Overview

3.5 Vocabularies

3.5.1 Controlled Vocabularies

3.5.1.1 Introduction

3.5.1.2 Indexing Schemes

3.5.1.3 Classification schemes

3.5.1.4 Thesauri

3.5.1.5 Taxonomies

3.5.2 Uncontrolled Vocabularies

3.5.2.1 Introduction

3.5.2.2 Folksonomies

3.6 Maps

3.6.1 Introduction

3.6.2 Semantic Nets

3.6.3 Frames

3.6.4 Mind Maps

3.6.5 Conceptual Graphs

3.6.6 Concept Maps

3.6.7 Neural Networks

3.6.8 Cognitive Maps

3.6.9 Fuzzy Cognitive Maps

3.6.9.1 Fuzzy Cognitive Maps

3.6.9.2 Rule-Based Fuzzy Cognitive Maps

3.6.9.3 Extended Fuzzy Cognitive Maps

3.6.9.4 Dynamic Cognitive Networks

3.6.9.5 Neural Cognitive Map

3.6.9.6 Neutrosophic Cognitive Maps

3.6.9.7 Probability Fuzzy Cognitive Maps

3.6.9.8 Fuzzy Relational Maps

3.6.10 Knowledge Maps

3.6.11 Topic Maps

3.6.12 Ontologies

3.6.12.1 Ontologies

3.6.12.2 Ontology Engineering

3.6.13 Other Knowledge Representation Structures

3.7 Analyzing Maps

3.7.1 Structural Comparison

3.7.2 Map Engineering

3.7.3 Mapping Tools

3.8 Harnessing Collective Intelligence for Knowledge Mapping

4. Data Mining the Semantic Web

4.1 Web Data Mining

4.2 Document Processing

4.3 The Case for Knowledge Representation Structures as Metadata

4.4. The Conceptual Knowledge Base

4.4.1 Introduction

4.4.2 Representational Elements of a Map

4.4.3 Operations with Maps

4.5 Mapping Descriptive Knowledge with Maps

4.5 Concepts in Documents’ classification

4.5.1 Introduction

4.5.2 Concept-Based Information Retrieval

4.5.3 Concept-Based Document Classification

5. Conclusions, Critics and Further Research

6 Appendix

6.1.Acronyms:

6.2 References: