A Gray code based ordering for documents on shelves: Classification for browsing and retrieval
read more
Citations
The Gray Code.
Constant-Weight Gray Codes for Local Rank Modulation
The structure of single-track Gray codes
Market basket analysis of library circulation data
Chip-integrated all-optical 4-bit Gray code generation based on silicon microring resonators.
Related Papers (5)
Frequently Asked Questions (12)
Q2. What is the difficulty of the method of analysis?
One difficulty associated with this method of analysis is that distances between documents are computed based solely on their subject headings.
Q3. What is the purpose of the proposed classification system?
The proposed classification system, consistent with a set of requirements for a classification system, can be used effectively to classify documents in library and database environments.
Q4. What are the five databases used for the proposed classification system?
The five databases used consist of bibliographic records retrieved from searches of the UNC-Chapel Hill Library’s on-line catalog.
Q5. What is the way to reduce the expected dissimilarity between features?
As the rightmost bits “cycle” more frequently than bits to the left, this arrangement will result in a lower expected dissimilarity between adjacent documents than would be the case if the least probable features with the greatest expected dissimilarity were on the right side and cycled most frequently.
Q6. What is the advantage of using a theoretically based model for a classification system?
One of the advantages of using a theoretically based model for a classification system is that classification performance can be analyzed.
Q7. How many binary features can be extracted from an existing classification system?
If the first character of an existing classification code has one of twenty six possible characters, the first twenty six binary feature values of the proposed system could be extracted and treated as a single character for comparison.
Q8. Why is the greater increase in performance due to the combined database?
The greater increase in performance for the combined database is due in part to the increase in , but is also due to the increased effectiveness of both classification systems when database size increases.
Q9. What is the effect of the dissimilarity function on a conceptual shelf?
This document arrangement has the effect of ordering documents so that at any one point on the conceptual shelf, the documents to both the right and the left of the point are arranged in order of increasing dissimilarity to the document at the chosen point.
Q10. What is the way to extend the search beyond subject searching?
The proposed system can incorporate both classification by subject and by incorporation of non-subject related bibliographic features, allowing for the extension of browsing beyond subject searching.
Q11. How can the expected dissimilarity between features be reduced?
the expected dissimilarity between documents, as represented by the sum of the expected dissimilarity between features, can be decreased by placing those features with the least expected dissimilarity furthest to the right in the code, while the features with the greatest expected dissimilarity are placed to the left.
Q12. What are the methods used to determine similarities between documents?
Methods similar to the work here include studies of coding techniques that place similar documents near each other, given a query [10], and hashing techniques that maintain order [13].