1
Кирил
no recommendations
no complaints
Closed
Project title: Help in writing a diploma on search in the DB
Type of cooperation: One-time project
Section: Education and consulting
Prepayment: without prepayment
Payment methods: Bank transfer
Acceptance of requests: closed
Type of cooperation: One-time project
Section: Education and consulting
Prepayment: without prepayment
Payment methods: Bank transfer
Acceptance of requests: closed
Project description:
There is a database that has a book table from the publishing house (ZBook) and a book table from the library (Description). The format of the book submission is bibliographic + there is is isbn and annotations.
So, the essence of the diploma:
Develop a method for identifying matching and similar entries in the library’s database of books and books from the publishing house. Develop criteria for assessing the effectiveness of the developed method and perform its experimental evaluation.
There are several options for the search result:
(1) The complete match between the book and the library.
(2) Semi-accuracy (Other edition, volume, year or place)
(3) Disparity. Here begins the search for similar records about the book.
If the records do not match, you need to issue a list of similar records, based on the title and annotation of the book. Not all books have an abstract, so you can ignore the records where there is no annotation. The method can be made very simple, the main thing is that it is not trivial, gives relevant results and works in an acceptable time (no more than 5 seconds). If you use full-text search for a mismatch, you need to link it to another method to improve the search or use another metric of similarity of records. It's a must.
For simplicity, the program can receive the entry ID books from the table of ordered books and display a list of similar books from the library.
The database of books from publishers is very small, since the hands did not reach the parser to write to extract data from the website of one of the publishers and fill the database with them. The library's book table has 400,000 entries, but actually there are fewer entries because there are still dissertations in it and there are a lot of old books in it.
The abstracts are presented for 60 thousand. Records.
Microsoft SQL Server 2014
You need to write in Python 3.5.
Similar literature can be displayed in the console. You can also infer the degree of similarity. If you have time, you should write a GUI for the program.
If you have any questions, write.
The price is negotiated.
Write reviews of methods and develop your method. Describe your method and agree with me. The method should produce relevant results. Develop criteria for assessing the effectiveness of the developed method and perform its experimental evaluation. Provide data on the criteria for assessing the effectiveness of the developed method and data on the experiments performed. Think of it as charts and tables. You can compare it with full text search MS sql server. Develop a working software that searches for similar records using the developed method. The search should preferably be carried out no more than 5 seconds. To speed up development, you can write a console version of the software, but if you have time, you can develop a GUI for the software. At the entrance, the program can receive book ID from the table of ordered books, at the output, the program displays a list of similar books that have coincided or found from the library with a similarity coefficient.
Microsoft SQL Server 2014
The operating system Windows 10.
There is a database that has a book table from the publishing house (ZBook) and a book table from the library (Description). The format of the book submission is bibliographic + there is is isbn and annotations.
So, the essence of the diploma:
Develop a method for identifying matching and similar entries in the library’s database of books and books from the publishing house. Develop criteria for assessing the effectiveness of the developed method and perform its experimental evaluation.
There are several options for the search result:
(1) The complete match between the book and the library.
(2) Semi-accuracy (Other edition, volume, year or place)
(3) Disparity. Here begins the search for similar records about the book.
If the records do not match, you need to issue a list of similar records, based on the title and annotation of the book. Not all books have an abstract, so you can ignore the records where there is no annotation. The method can be made very simple, the main thing is that it is not trivial, gives relevant results and works in an acceptable time (no more than 5 seconds). If you use full-text search for a mismatch, you need to link it to another method to improve the search or use another metric of similarity of records. It's a must.
For simplicity, the program can receive the entry ID books from the table of ordered books and display a list of similar books from the library.
The database of books from publishers is very small, since the hands did not reach the parser to write to extract data from the website of one of the publishers and fill the database with them. The library's book table has 400,000 entries, but actually there are fewer entries because there are still dissertations in it and there are a lot of old books in it.
The abstracts are presented for 60 thousand. Records.
Microsoft SQL Server 2014
You need to write in Python 3.5.
Similar literature can be displayed in the console. You can also infer the degree of similarity. If you have time, you should write a GUI for the program.
If you have any questions, write.
The price is negotiated.
Write reviews of methods and develop your method. Describe your method and agree with me. The method should produce relevant results. Develop criteria for assessing the effectiveness of the developed method and perform its experimental evaluation. Provide data on the criteria for assessing the effectiveness of the developed method and data on the experiments performed. Think of it as charts and tables. You can compare it with full text search MS sql server. Develop a working software that searches for similar records using the developed method. The search should preferably be carried out no more than 5 seconds. To speed up development, you can write a console version of the software, but if you have time, you can develop a GUI for the software. At the entrance, the program can receive book ID from the table of ordered books, at the output, the program displays a list of similar books that have coincided or found from the library with a similarity coefficient.
Microsoft SQL Server 2014
The operating system Windows 10.
