book book id_criteria - freebase-schema/freebase GitHub Wiki

• title_match: There is a title_match if the titles of the books on both the sources are same or match approximately. Translated titles or localized titles are considered matches. Harry Potter and the Philospher's Stone and Harry Potter and the Sorcerer's Stone are a match. One Hundred Years of Solitude and Cien años de soledad are a match.

• author_match: There is an author_match if at least one of the book authors listed under Freebase property /book/written_work/author (Author) matches with the authors on the other source.

• editions_match: There is an editions_match if at least one of the book editions (with edition title, year, publisher) listed under freebase property /book/book/editions (Editions) or /book/book/first_edition (First edition) matches with the editions on the other source.

• first_publication_date_match: There is a first_publication_date_match if the date (to the day level) that the book was first published in any form (i.e., hardcover, softcover, e-book) listed under freebase property /book/written_work/date_of_first_publication (Date of first publication) is same as the date on the other source. Sometimes only the year of publication may be available and if it matches on both sources, it is a match.

• part_of_series_match: There is a part_of_series_match if this book is part of a series and the series listed under the Freebase property /book/written_work/part_of_series (Part of series) matches with the series on the other source.

• cover_page_match:There is a cover_page_match if the illustrated cover page of the book on freebase topic is same as the cover page illustration on the other source.

• awards_match: There is an awards_match if at least one of the nominations or awards (country/association, category and year) received by the book matches with the nominations or awards on the other source.

Books should not be matched to other incompatible types such as book edition, TV program, or film (Example 8).

Here are some typical patterns for determining identity of two books:

1. title_match and author_match (Example 1)
2. title_match and first_publication_date_match (Example 2)
3. title_match and part_of_series_match (Example 3)
4. title_match and editions_match (Example 4 , Example 5)
5. title_match and awards_match (Example 6)
6. title_match and cover_page_match (Example 7)

Pattern 1 Example 1: Permalink Here the title of the books and their authors match on both the sources. Hence the books are identical.

Pattern 2 Example 2: Permalink In this example the title of the books and the full date of their first release match on both the sources. Hence the books are identical.

Pattern 3 Example 3: Permalink Here the title of the books match. Also the name of the series that the book is a part of, match. Hence the books are identical.

Pattern 4 Example 4: Permalink The title of the books and details about one of the editions (Published Year, Publisher) of the book match on both the sources. Hence the books are identical.

Pattern 4 Example 5: Permalink The title of the books and the information about first edition of the book (first edition title, publishing year, and publisher) from Left Hand Side match with the information on Right Hand Side. Hence the books are identical.

Pattern 5 Example 6: Permalink The title of the books and the awards received by the books match on both the sources. Hence the books are identical.

Pattern 6 Example 7: Permalink The title of the books and their cover pages are an exact match. Hence the books are identical.

Example 8: Permalink Here the topic titles match on both the sources, but while the freebase topic on Left Hand Side is about the book, the Right Hand Side talks about a particular edition of the book published in 2009 (the first edition being published in 1983). Hence these two topics should not be reconciled.