Create a prediction

  • serials
Last updated: 10-08-2023


Predictions are a field of serial holdings and can streamline the management of scheduled serials by defining an automatic generation and display of future issues according to the frequency of publication.

Template Defines the model for the numbering display format, using the names of prediction patterns and chronology levels. These names are variables to be defined in the prediction patterns. Example : {chronology.month} {chronology.year}.
Three preset variables can be used: {}, {expected_date.month} et {expected_date.year}. They are used to display the day, month or year of the issue in the model, according to the defined frequency.
Periodicity Indicates the default publication frequency for calculating the issue's numbering.
Expected date for the next issue Defines the date on which the system expects to receive the next serial issue. If it is not received, it will be marked as "late".
Prediction pattern Pattern comprising one or more chronology level(s) that are interdependent in their incrementation
Chronology level Variable to be incremented according to predefined parameters. It can be numeric (1, 2, 3, ...) or based on a predefined list (months, seasons, ...).
Starting value* Defines the start of the sequence. Example: 1961 for a magazine first published in that year.
Completion of the values* Defines the end of the sequence. For an upper chronology level (at the very top of the prediction pattern), nothing is generally entered; for example, for the year, we don't know the end year of a periodical. For a lower chronology level, the upper level is incremented by one when this value is reached, and the lower level sequence starts again from the starting value.
Next predicted value* Value for the next expected issue.

* For a number based chronology level, the value to be entered is the desired number. Otherwise, it's the number of one of the values specified in the list.

Several independent patterns can be defined, each with one or more chronology levels to correspond to the specific numbering of the catalogued document.

Example of a prediction pattern

Let's take the simple example of a monthly magazine. The issues of such a magazine follow the pattern: No 492, July 2021; No 493, August 2021; No 494, September 2021; ....

This magazine's holding needs two prediction patterns incremented independently (one incrementation has no influence whatsoever on the other):

  • the numbering: 37, 38, 39, etc.
  • chronological indication: December 2019, January 2020, February 2020, etc.

In the editor, subfields can be created as follows:

  • Prediction pattern 1: Defines the numbering -- Name: numbering
    • Chronology level (Number based prediction): calculates the increment of the issue number -- Name: no
  • Prediction pattern 2: defines the chronology (month, year) -- Name: chronology
    • Chronology level 1 (List based prediction (used for seasons, months...): calculates month increment -- Name: month
    • No need to define the year: the predefined variable {{expected_date.year}} can be used, as the year is displayed in digits, unlike the months, which in this case must be displayed in full words.

Settings and template

In the Template field, we enter the numbering as it should be displayed and we use the names of the chronology levels between {{ }} for incrementable values that must be calculated for each issue: No {{}}, {{chronology.month}} {{expected_date.year}}

Prediction pattern

The Name fields are are always free and arbitrary. These variables are used between {{ }} in the model.

In our example, the magazine is not published every month. That's why we've defined a list based prediction in chronology level 2. This list represents a standard publication yearly sequence. In our case, it was sufficient to combine the months of July and August into a single list item. This modularity makes it possible to predict more irregular sequences, or sequences based on other units (seasons, etc.).

Periodicity change

When a serial changes periodicity, two options are available:

  1. Replace the current prediction with the new one. Make sure you have received all the expected issues for the previous frequency, otherwise they will have to be added as irregular issues.
  2. Create a new holdings with a new prediction.

Create a serial holdings | Manage serial issues