Hi, Data migration can be a very involved process for large datasets. For small datasets, you are better off using MARCEdit with the process described in your article. That is, you will have to place each item barcode with their respective details in separate row, then converting it to MARC with item details in tag 852 (Koha uses tag 952 with few extra sub-fields). Once the MARC data is loaded into the library system, you will have to move all 'copy' item records into single bib record manually. Then you may delete all bib records without items, which can be done in batch mode. However, if your dataset is large (say more than 100,000) manual approach is not recommended. It will be prudent to get the migration done by people who can write custom code to consolidate the item records into their respective bib records before loading it into the system. There could be many approaches to it, we normally use the following steps in large datasets: 1. Prepare the data: make it consistent across all columns (date format, capitalization, white-spacing, call number/book number, etc.). In the absence of it, lot of cleaning will have to done either during MARC conversion through custom code or manually after data load. 2. Sort the records by title, if possible. It can also be done later using MARCEdit or a custom script after MARC conversion. 3. Convert the records to MARC (either through MARCEdit or custom code). We prefer custom code as it gives more flexibility. 4. Consolidate item records to single bib: I'm afraid there is no ready made tool for this. A custom program is required to loop through the MARC records (which is created in the previous step, sorted by title). In this step we match current record's title, call number (optionally edition, if available) with the previous record in each iteration. If the match is found, then only the item tag (952, in case of Koha) is copied and appended to the previous MARC record (and rest of current record ignored). This program's output will create the final MARC record set that will have multiple copies populated in repeatable item tag for each bib record. MARC data manipulation modules/libraries are available in most programming languages. For someone with intermediate programming skills, a bit of effort will get you far better result with custom program than easy-to-use tools like MARCEdit especially for large datasets. Best regards, Saiful Amin ​Semantic Consulting, Bangalore​ +91-9343826438 On Tue, Jun 20, 2017 at 10:37 AM, basudeb adhikary < basudebadhikary@rediffmail.com> wrote:

Dear Sir,

In response to your enriching "data migration videos", I am writing the mail. Surely these are very much knowledge enriching video. Actually we had undertaken a similar type of job in our institute and the experience was shared in http://www.ijidt.com/index.php/ijidt/article/view/306. Now I have a question. What about the 852$p - the accession number field? Say, I have a title having 5 copies having different accession numbers. How should I arrange them in Excel sheet so that they get added in the software specifically SOUL 2.0. I know that 852 in the local field but this is required for migrating internal database only. I will be highly grateful to you if you kindly show us some light in this direction.

Warm Regards Basudeb Adhikary Librarian Netaji Mahavidyalaya Hooghly West Bengal