Script Recognition with Hierarchical Feature Maps

@article{Miikkulainen1992ScriptRW,
  title={Script Recognition with Hierarchical Feature Maps},
  author={Risto Miikkulainen},
  journal={Connection Science},
  year={1992},
  volume={2},
  pages={196-214}
}
The hierarchical feature map system recognizes an input story as an instance of a particular script by classifying it at three levels: scripts, tracks and role bindings. The recognition taxonomy, i.e. the breakdown of each script into the tracks and roles, is extracted automatically and independently for each script from examples of script instantiations in an unsupervised self-organizing process. The process resembles human learning in that the differentiation of the most frequently… 

Exploration of text collections with hierarchical feature maps

TLDR
This paper shows the results from using a hierarchy of self-organizing maps to perform the text classification task, and the choice of this particular artificial neural network model enables the establishment of a document taxonomy.

Script-based inference and memory retrieval in subsymbolic story processing

TLDR
DISCERN shows that high-level natural language processing is feasible through integrated subsymbolic systems and learns to infer unmentioned events and unspecified role fillers, generates expectations and defaults, and exhibits plausible lexical access errors and memory interference behavior.

Automatic labeling of self-organizing maps for information retrieval

  • D. MerklA. Rauber
  • Computer Science
    ICONIP'99. ANZIIS'99 & ANNES'99 & ACNN'99. 6th International Conference on Neural Information Processing. Proceedings (Cat. No.99EX378)
  • 1999
TLDR
This work presents the novel LabelSOM method, which automatically selects the most descriptive features of the input patterns mapped onto a particular unit of the map, thus making the characteristics of the various clusters within the map explicit.

Classification with Unsupervised Artificial Neural Networks

TLDR
It is shown that a similar display can be obtained by using hierarchical feature maps to represent the contents of a document archive while still having the general maps available for global orientation.

Automatic Labeling of Self-Organizing Maps forInformation RetrievalDieter Merkl

TLDR
This paper presents the novel LabelSOM method, which automatically selects the most descriptive features of the input patterns mapped onto a particular unit of the map, thus making the characteristics of the various clusters within the map explicit.

CIA's View of the World and What Neural Networks Learn from It: A Comparison of Geographical Document Space Representation Metaphors

TLDR
It is shown that a similar display can be obtained by using hierarchical feature maps to represent the contents of a document archive by starting from a map representing the complete data collection different regions are shown at finer levels of granularity.

Text Mining in the SOMLib Digital Library System: The Representation of Topics and Genres

TLDR
The SOMLIb digital library system, built on neural networks to provide text mining capabilities, is presented and the capabilities of the SOMLib system are demonstrated using collections of articles from various newspapers and magazines.

Content-Based Image Retrieval Combining Growing Hierarchical Self-Organizing Maps and a Log-Polar Representation *

TLDR
This paper presents an approach for content-based image retrieval, which combines GH-SOM (Growing Hierarchical Self-Organizing Maps) and a log-polar representation, which has a property that helps achieve some degree of scale and orientation invariance.
...

References

SHOWING 1-10 OF 36 REFERENCES

A modular neural network architecture for sequential paraphrasing of script-based stories

TLDR
Sequential recurrent neural networks have been applied to a fairly high-level cognitive task, i.e. paraphrasing script-based stories, using hierarchically organized modular subnetworks and is able to produce a fully expanded paraphrase of the story from only a few sentences.

Scripts in memory for text

Mixing Binary and Continuous Connection Schemes for Knowledge Access

TLDR
The problems of combining two types of connection system are discussed and a new tool for connection systems, the pulse-out, which replaces the Boltzmann Machine in creating energy leaps is introduced.

Mechanisms of Sentence Processing: Assigning Roles to Constituents of Sentences

TLDR
A simulation model is described that exhibits considerable facility in dealing with the problems of frame selection, role assignment, disambiguation, etc, and suggests a natural way to resolve unappealing aspects of the idea that there is a fixed set of individuated case roles.

Reconstructive Memory: A Computer Model

TLDR
A conceptual memory organization implied by human reconstructive memory is presented along with examples which motivate it and conclusions are drawn concerning retrieval failures and the relationship of episodic and semantic memory.

Scripts, plans, goals and understanding: an inquiry into human knowledge structures

For both people and machines, each in their own way, there is a serious problem in common of making sense out of what they hear, see, or are told about the world. The conceptual apparatus necessary

A framework for representing knowledge