Part I--A Prototype System for Data Exploration |
|
1 | (132) |
|
1 The Data to be Explored |
|
|
2 | (10) |
|
1.1 Different Types of Phenomenological Data |
|
|
3 | (6) |
|
|
4 | (3) |
|
|
7 | (2) |
|
|
9 | (2) |
|
|
11 | (1) |
|
|
12 | (16) |
|
2.1 Identifying The Problem |
|
|
12 | (3) |
|
2.1.1 Heterogeneous Data Models |
|
|
13 | (1) |
|
2.1.2 Data Assumptions of Algorithms |
|
|
14 | (1) |
|
2.2 The Concept of Data Exploration |
|
|
15 | (1) |
|
2.2.1 Purpose of Data Exploration |
|
|
15 | (1) |
|
|
16 | (1) |
|
|
16 | (6) |
|
2.3.1 Computational Goals |
|
|
17 | (1) |
|
2.3.2 Task-related Requirements |
|
|
18 | (2) |
|
2.3.3 Data Exploration for International Relations Research |
|
|
20 | (2) |
|
2.4 Comparison to Other Approaches |
|
|
22 | (6) |
|
2.4.1 Systems Which Implement a Domain-Theory |
|
|
22 | (1) |
|
|
23 | (3) |
|
2.4.3 Information Retrieval and Hypertext |
|
|
26 | (2) |
|
|
28 | (56) |
|
3.1 Object-oriented Data Exploration |
|
|
28 | (7) |
|
3.1.1 Reasons for An Object-oriented Approach |
|
|
29 | (3) |
|
3.1.2 CrisView's Main Components |
|
|
32 | (2) |
|
3.1.3 The SHERFACS Dataset |
|
|
34 | (1) |
|
|
35 | (9) |
|
|
36 | (1) |
|
3.2.2 Special Data Interface Operators |
|
|
37 | (4) |
|
3.2.3 Characteristics of the Document Concept |
|
|
41 | (2) |
|
3.2.4 Documents in An Object-oriented System |
|
|
43 | (1) |
|
|
44 | (9) |
|
|
44 | (2) |
|
|
46 | (5) |
|
3.3.3 Acceleration Structures |
|
|
51 | (1) |
|
3.3.4 Reinforcing the Coding Schema |
|
|
52 | (1) |
|
|
53 | (30) |
|
3.4.1 Exploration Techniques |
|
|
54 | (2) |
|
|
56 | (3) |
|
|
59 | (3) |
|
3.4.4 New-variable-window |
|
|
62 | (3) |
|
|
65 | (2) |
|
|
67 | (3) |
|
|
70 | (5) |
|
3.4.8 Rule-analysis-window |
|
|
75 | (3) |
|
|
78 | (5) |
|
|
83 | (1) |
|
4 Knowledge-based Analyses Techniques |
|
|
84 | (45) |
|
4.1 Selection of Knowledge-based Techniques |
|
|
85 | (1) |
|
|
85 | (25) |
|
|
87 | (2) |
|
4.2.2 Algorithm Generalizations |
|
|
89 | (5) |
|
4.2.3 Entropy-based Decision-Tree Induction |
|
|
94 | (1) |
|
|
95 | (6) |
|
|
101 | (7) |
|
4.2.6 Object-oriented Induction |
|
|
108 | (2) |
|
4.3 Precedent-based Reasoning |
|
|
110 | (14) |
|
|
111 | (7) |
|
4.3.2 Retrieving Precedents |
|
|
118 | (3) |
|
4.3.3 Precedent-based Reasoning and Induction |
|
|
121 | (3) |
|
4.4 Problems of Analyzing Real-world Datasets |
|
|
124 | (4) |
|
4.4.1 Divergence from the Coding Scheme |
|
|
125 | (1) |
|
4.4.2 The Heterogeneity of Datasets |
|
|
125 | (1) |
|
4.4.3 Methodological Issues |
|
|
126 | (2) |
|
|
128 | (1) |
|
5 Summary and Conclusion to Part I |
|
|
129 | (4) |
|
|
129 | (2) |
|
5.2 Leading Over to Part II |
|
|
131 | (2) |
Part II--A Protocol for Data Exploration Systems |
|
133 | (70) |
|
|
135 | (4) |
|
6.1 Providing Assistance to Users |
|
|
136 | (1) |
|
|
137 | (2) |
|
|
139 | (9) |
|
|
140 | (1) |
|
7.2 Drawing Analogies to Data Exploration Systems |
|
|
141 | (1) |
|
7.3 A Protocol for Data Exploration Systems |
|
|
142 | (6) |
|
7.3.1 Three Additional Protocol Components |
|
|
142 | (2) |
|
7.3.2 Steps Toward a Dedicated Application Framework |
|
|
144 | (1) |
|
7.3.3 Development of Components |
|
|
145 | (3) |
|
8 Extending the CrisView Protocol |
|
|
148 | (45) |
|
8.1 Developing Base Classes |
|
|
148 | (27) |
|
8.1.1 Mechanism of Data Access and Data Analysis |
|
|
149 | (3) |
|
8.1.2 Generalizing the Import Behavior |
|
|
152 | (13) |
|
8.1.3 Constructing Acceleration Structures |
|
|
165 | (10) |
|
8.2 Developing Specialized Classes |
|
|
175 | (10) |
|
8.2.1 Specialized Classes for Representing Data |
|
|
176 | (4) |
|
8.2.2 Specialized Classes for New Data Exploration Tasks |
|
|
180 | (5) |
|
8.3 Developing Introspective Classes |
|
|
185 | (5) |
|
8.3.1 Introspective Access for End Users |
|
|
186 | (2) |
|
8.3.2 Introspective Access for Specific User Groups |
|
|
188 | (2) |
|
8.4 Generalizing the User Interface |
|
|
190 | (3) |
|
|
193 | (10) |
|
9.1 Evaluation of CrisView |
|
|
194 | (4) |
|
9.2 The Use of Metaobject Programming |
|
|
198 | (5) |
References |
|
203 | (14) |
Index |
|
217 | |