Automated aggregation of geographic objects. A new approach to the conceptual generalisation of geographic databases
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Automating the process of map generalisation has been a scientific challenge for over 30 years and still there is no comprehensive practical method.In this study I deal with the database part of the problem which has received relatively little attention in comparison with the total effort put into map generalisation research.This way I hope to contribute to a generic map generalisation system that should be possible before long, by combining the research results of the past few decades.This study was carried out during a few distinct periods.I first got involved in the subject of map generalisation in 1992 at the Canada Centre for Remote Sensing in Ottawa, where I assisted Dianne Richardson in implementing the method she had developed for her PhD research.A period that I look back upon with great pleasure.Subsequently, in 1994, I started a four-year PhD research project at Wageningen University, which was a followup of Dianne's work.It was during this time that I developed the concept of aggregation based on co-occurrence of classes.After this period the project came to a temporary standstill.Although I never lost the intention to finish it, it was only last year that I picked it up again and finished it, resulting in this dissertation.During these periods a number of people have been involved.I would like to thank Martien Molenaar for his patience when it might have seemed that I would not finish my study with a dissertation, for providing me with the conceptual framework for my study and for leaving me the freedom to find my own solutions.Arnold Bregt, who got involved in the project at a later stage, but still provided invaluable input and very practical comments when most needed.As during one meeting at ITC, when we solved the remaining issues in a very constructive 15-minute discussion.I would further like to thank all colleagues at the Surveying department (Landmeetkunde) for their collegiality and countless gezellige lunches at Unitas on Tuesdays when there were pannekoeken on the menu.Special thanks go to John Stuiver for introducing me to GIS in the first place and making me enthusiastic about its possibilities.Ren van der Schans for inspiring and animated discussions during the early stages of my study.Ron van Lammeren for all his help and enthusiasm.Elisabeth Addink, of course, for her wit and our collegial discussions, occasionally on the subject but most of the time diverting into the most ridiculous directions, which also made the time at Landmeetkunde memorable.Further thanks go to Kees Bol for salvaging files that I had accidentally deleted from the server and Prof. Kruidhof, founder of the Landmeetkunde laboratory, for providing us with such an exceptional place to work, overlooking the ever-changing sight of the river Rhine flood plains.Lots of thanks go to my parents, who have always encouraged me to study.Well mum, dad, this is about all I can do.And finally my girlfriend Ingrid for her love and support, as well as patiently answering people's questions whether I had finished my PhD yet.The goal of this study is to develop a framework and a working prototype for the generalisation of object-and vector-based categorical maps -such as large-scale topographic data -based on inter-object relationships.We strive for a system that is to a large extent automated and can be operated by non-expert users.Large-scale topographic maps are commonly object-based, categorical maps.The objects are classified: 'road', 'building' etc.The spatial inter-object relationships in large-scale datasets are often complicated.We will concentrate on the aggregation of objects belonging to distinct object classes, based on the spatial and thematic relationships
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.002 |
| Open science | 0.002 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it