Information storage and retrieval

An information storage system in which information items are processed so as to map to respective nodes in an array of nodes by mutual similarity of the information items, such that similar information items map to nodes at similar positions in the array of nodes to form a self-organizing map, the system comprising:

• a processor;
• a generator configures to generate a feature vector derived from each information item of the self-organizing map, the feature vector for an information item representing a set of frequencies of occurrence, within that information item, of each of a group of information features;
• mapping logic configured to map each feature vector to a node in the self-organizing map, the mapping between information items and nodes in the array including a dither component configured to display nodes that have substantially identical or identical information items at different locations arising from an application of the dither component in a display area to visibly distinguish the nodes having substantially identical or identical information items;
• logic configured to mare a newly received information item to a node in the array of nodes;
• a mapping error detector configured to detect a mapping error as the newly received information items is so mapped;
• and
• logic, responsive to a detection that the mapping error exceeds a threshold error amount, configured to initiate a remapping process of the set of information items and the newly received information item,
• wherein the dither component is a random addition to a node position of up to a half of separation distance between adjacent nodes.

An information storage method in which information items are processed so as to map to respective nodes in an array of nodes by mutual similarity of the information items, such that similar information items map to nodes at similar positions in the array of nodes to form a self-organizing map, the method comprising:

• generating a feature vector derived from each information item of the self-organizing map, the feature vector for an information item representing a set of frequencies of occurrence, within that information item, of each of a group of information features;
• mapping each feature vector to a node in the self-organizing map, the mapping between information items and nodes in the array including a dither component configured to display nodes that have substantially identical or identical information items at different locations arising from an application of the dither component in a display area to visibly distinguish the nodes having substantially identical or identical information items;
• mapping a newly received information item to a node in the array of nodes;
• detecting a mapping error as the newly received information items is so mapped;
• and
• initiating a remapping process of the set of information items and the newly received information item logic, in response to a detection that the mapping error exceeds a threshold amount,
• wherein the dither component is a random addition to a node position of up to a half of separation distance between adjacent nodes.

An information retrieval method in which a set of distinct information items map to respective nodes in an array of nodes by mutual similarity of said information items, so similar information items map to nodes at similar positions in said array of nodes to form a self-organizing map, said method comprising:

• (i) displaying a representation of at least some of said nodes of the self- organizing map as a two-dimensional display array of display points within a display area;
• (ii) defining, with a user control, a two-dimensional region of said display area;
• (iii) detecting those display points lying within said two-dimensional region of said display area;
• (iv) displaying, concurrently with the representation of at least some of said nodes, a list of data representing information items, being those information items mapped onto said nodes corresponding to display points displayed within said two-dimensional region of said display area;
• (v) computing a quantization error of a newly received information item, comparing the quantization error to the self-organizing map, retraining the self-organizing map when the quantization error is above a predetermined threshold;
• and
• (vi) applying a dither component so as to display nodes that have substantially identical or identical information items at different locations in a display area, to visibly distinguish the nodes have substantially identical or identical information items, the dither component being a random addition to a node position of to a half of a separation between adjacent nodes.