Method for retrieving high relevance, high quality objects from an overall source

Claim 3

The method according to claim 1, further comprising a step (4) of determining dominance relations between every two identified objects having directed relations based upon a comparison of their relevance probabilities to each other, and ordering the identified objects in accordance with their dominance relations.

Claim 4

The method according to claim 2, further comprising a step (5) of ordering the identified objects in accordance with their dominance relations as a digraph tree, and identifying the roots and branch points of the digraph tree as an ordering of dominant objects.

Claim 7

The method according to claim 1, further comprising a step (4) of determining dominance relations between every two identified journals having citation relations based upon a comparison of their relevance probabilities to each other, and ordering the identified journals in accordance with their dominance relations.

Claim 8

The method according to claim 5, further comprising a step (5) of ordering the identified journals in accordance with their dominance relations as a digraph tree, and identifying the roots and branch points of the digraph tree as an ordering of dominant journals.

Claim 9

The method according to claim 6, wherein the roots and branch points of the digraph tree are ordered in accordance with their dominance relations as a condensation graph.

Claim 10

The method according to claim 4, wherein the threshold relevance value is selected at a maximal value representing the highest probability of relevance of another journal of the source collection with the starter journal.

Claim 11

The method according to claim 4, wherein journals having citation relations with each other form a bi-communication class of journals, and the threshold relevance value is selected at a lowest relevance probability value towards 0 at which a last bi-communication class of journals can be identified from the source collection of journals.

Claim 12

The method according to claim 1, wherein the objects are published papers that have bibliographic references to other published papers of the source collection of published papers, and said method is applied for retrieval of high-relevance published papers from said source collection.

Claim 13

A computer system, including program means and a database storage, for selecting a core collection of journals from an overall source collection, wherein the journals have directed relations with each other in the form of citation of other journals by a journal, and the numbers of citations of other journals by each journal are given as citation number data of the database, comprising: first program means for computing, from the citation number data, a highest number of citations of a journal by other journals of the source collection and selecting it as a starter journal; second program means for computing, from the citation number data, frequencies of citation of other journals of the source collection from the starter journal as measures of probabilities of relevance of the other journals to the starter journal; third program means for inputting a selection of a threshold relevance value, for identifying all journals of the source collection having relevance probabilities to the starter journal exceeding the threshold relevance value, and for re-executing said determining of a starter journal by said first program means and said computing of the probabilities of relevance by said second program means for each identified journal until no new journals are identified, and output means for providing a listing of the identified journals as a core collection of high-relevance journals selected from the overall source collection.

Claim 14

A computer system, including program means and a database storage, for selecting a core collection of journals from an overall source collection, wherein the journals have directed relations with each other in the form of citation of other journals by a journal, and the numbers of citations of other journals by each journal are given as citation number data of the database, comprising: first program means for computing, from the citation number data, a highest number of citations of a journal by other journals of the source collection and selecting it as a starter journal; second program means for computing, from the citation number data, frequencies of citation of other journals of the source collection from the starter journal as measures of probabilities of relevance of the other journals to the starter journal; third program means for inputting a selection of a threshold relevance value, for identifying all journals of the source collection having relevance probabilities to the starter journal exceeding the threshold relevance value, and for re-executing said determining of a starter journal by said first program means and said computing of the probabilities of relevance by said second program means for each identified journal until no new journals are identified, and output means for providing a listing of the identified journals as a core collection of high-relevance journals selected from the overall source collection.

Claim 15

The system according to claim 11, further comprising fourth program means for determining dominance relations between every two identified journals having citation relations based upon a comparison of their relevance probabilities to each other, and for ordering the identified journals in accordance with their dominance relations.

Claim 16

The system according to claim 12, further comprising fifth program means for ordering the identified journals in accordance with their dominance relations as a digraph tree, and for identifying the roots and branch points of the digraph tree as an ordering of dominant journals.