Acquisition management system

Up
We have developed an acquisition management system using Wrike for the coordination of acquisition among workers. The need for this arises from the possibility of wasted effort by workers inadvertently doing duplicate work, and also from the opportunity for workers to help each other.

The system is based on subject-language targeting. It presumes that the workers are starting their discovery efforts by choosing specific languages to discover data on. The system defines high-priority languages, based on our acquisition priorities, so you don’t need to use the PanLem priority-language selector. The system lets you choose among those languages. It requires you to report the choices that you make and your discovery progress, so other workers know that your chosen languages are being (or have been) worked on and what the results have been. This permits other workers to avoid choosing the same languages. But, if they happen to discover data on one of your languages, they can let you know. Finally, the system also manages procurement.

Set up a Wrike account

  1. Click “Accept invitation” in the email message you received from Wrike. This will open a Wrike screen.
  2. On the Wrike screen, enter your email and create a password. Click “Save Changes”.
  3. At upper right, click your name > Profile to upload a photo and edit your first and last name if necessary. Click “Save Changes”.

Display the Acquisition homepage

  1. Click the Wrike logo at upper left or click your name at upper right and choose “Go to Workspace”.
  2. Click the “Acquisition Languages” folder in the far left menu. A list of languages will display at the right.
  3. Click the arrow next to “Acquisition Languages” in the far left menu. A list of countries will display below.
  4. Click “Table” near the top of the screen.
  5. Hover/drag the arrow between column headings to adjust column widths.

The screen displayed here will be our Acquisition homepage. As you gain more experience with Wrike, you may find other views useful, such as List view.

Peruse languages by country, family, or importance

By country

  1. On the Acquisition homepage, click a country name at the far left, or the plus sign next to the country name in the table. You will see a list of the languages in that country for which PanLex seeks lexical translations.
  2. To go back to displaying all countries, click the “Acquisition Languages” folder at the far left.

By family or subgroup

  1. On the Acquisition homepage, click the small filter icon near the upper left corner of the table.
  2. At far right, click Advanced Filters > Custom Fields, then expand the Custom Fields list.
  3. Click on the desired filters and enter the item you are filtering for, spelled exactly as it appears in the Wrike table.

By importance

Follow the same steps as for family or subgroup. All of the languages on the Wrike list are needed in the PanLex database. The default importance level is “Normal”. A language is marked “High” importance if it belongs to a family that has fewer than 5,000 total expressions and fewer than 10 total sources in the database.

Choose a language to work on

  1. When you find a language to research, hover the mouse over the cell containing the language name and click the arrow that appears on the right side of the cell. This opens up a detailed view of the language task. The task description contains a list of alternative names for the language and links to its Glottolog and Ethnologue pages (if available).
  2. Change the language task status from “Open” to “In Progress”.
  3. Assign the language task to yourself and any partners.
  4. Start looking for sources on the language.
  5. Put any general information about your search for the language (i.e., whatever is not specific to a particular source) in the description and comments of the language task.

Add a new source subtask

Sources are represented in Wrike as subtasks of languages. There should be exactly one source subtask for every PanLex source. When you find a possible resource for a language, you need to create a subtask for the new PanLex source(s):

  1. Open the list of subtasks in the language task view by clicking the “Add subtask” button. (It may simply read “0” if your window is too small.)
  2. Enter a title for the new source subtask. This is typically the resource name but can be anything you like, as long as it allows easy identification of the source at a glance. (The subtask title will not be recorded in the PanLex database.)
  3. Hit return to create the subtask. The PanLex Bot will automatically add it to the “Acquisition Sources” folder and generate the standard source template in the subtask description. (This should happen promptly, but in some cases you may need to close and reopen the subtask view to see the template.)
  4. Open the subtask you created by clicking on its title.
  5. Change the status of the new subtask from “Open” to “In Progress”.
  6. Assign the subtask to yourself and any partners.

Repeat the above steps for each source you find for a language.

If you find a resource with multiple sources, you can fill out one source and use it as a template for the other sources. To do so, right-click the source subtask and select Duplicate. For clarity, you should give each source subtask that belongs to the same resource a distinct title. Note that once you have duplicated a source subtask, any changes you make to one subtask will not affect the other.

Enter source data

Bibliographic and other information about the source should be entered into the description of the Wrike subtask. See the source registration guidelines for more information on how to fill out the fields. Be careful not to change any field names or other formatting; otherwise the data may not be correctly processed. Miscellaneous information that is useful during the acquisition process (but will not end up in the database) can be stored at the bottom of the description under “NOTES”. It’s best if you put this information on the lines after “NOTES” and not on the same line.

Associate a source with multiple languages

Sometimes you may find a source that contains translations for more than one language in the “Acquisition Languages” folder on Wrike. In these cases, you should use a single source subtask that is associated with multiple languages, not multiple source tasks. As you fill out the list of language varieties in the source subtask description, the PanLex Bot will automatically associate the subtask with the appropriate language tasks (about 8 minutes after you finish making changes).

Teamwork

Please liberally use the notes and comments sections of your Wrike task screens to save documents and PDFs, discuss sources, leave comments, and support each other in determining if a source is appropriate for PanLex. Change the subtask status to “Review” to invite comments on a source.

Task status

We have created a number of custom statuses for language and source tasks in the “Acquisition” workflow on Wrike:

  • Open: indicates that a language or source is unassigned and available to be worked on.
  • In Progress: indicates that a language or source is assigned and being worked on.
  • Consult in Library: indicates that a resource is available in a local library and should be consulted.
  • Request from Library: indicates that a resource is available in a local library and should be requested (e.g. from a remote storage facility) so it can be consulted.
  • High-hanging Fruit: indicates that work on a language or source has stalled because it is difficult or challenging in some way.
  • Review: indicates that a source task is ready for review by the Source Acquisition Specialist or another worker.
  • Order: indicates that review has been approved and a non-digital resource should be ordered.
  • Scan: indicates that a portion of the publication must be scanned either through the UCB Library or by PanLex staff because item is out of print or only a few pages are needed.
  • Upload to Resource Archive: indicates that review has been approved and a digital resource should be uploaded to the resource archive. After the upload, the PanLex Bot will post a comment on the task confirming it.
  • Dead End: indicates that work on a language or source has ceased.
  • Error: indicates that the PanLex Bot has reported an error.
  • Completed: indicates that work on a language or source has been completed. For a source, also indicates that it has been registered in the PanLex database.

Custom fields

We have defined certain custom fields for language and source tasks. Custom fields are visible as columns in Table view, or by opening the custom field tab in the task detail view.

For language tasks, the available fields are:

  • Importance: automatically populated from the task importance, which you can set by clicking on the exclamation mark in the task detail view. This facilitates filtering and sorting tasks by importance.
  • Family: the language’s family, according to Glottolog.
  • Subgroup: the language’s immediate parent subgroup, according to Glottolog.
  • ISO 639-3 Code: automatically populated from the task title. This facilitates filtering and sorting tasks by ISO 639-3 Code.
  • Glottocode: already set for most language tasks, but not automatically set for language tasks created by regular (non-bot) users.
  • PanLex UID: automatically populated from the task title. This facilitates filtering and sorting tasks by PanLex UID.

For source tasks, the available fields are:

  • Approved: checkbox indicating whether a source has been approved for purchase and/or uploading.
  • Digital: field indicating whether a source is digital. Automatically populated based on the formats entered in the task description. Possible values are “yes”, “no”, and “unknown”. (The value is “unknown” when no formats, or invalid formats, have been entered.)
  • Supplier Name: automatically populated from the the task description.
  • Date Ordered: automatically populated from the task description.
  • Date Rec’d: automatically populated from the task description.
  • Amt Paid: automatically populated from the task description.
  • Amt Reimb’d: automatically populated from the task description.
  • IA Batch Num: automatically populated from the task description.
  • Importance: automatically populated from the task importance, which you can set by clicking on the exclamation mark in the task detail view. This field was added to facilitate filtering and sorting tasks by importance.

Complete a source

Once a digital resource has been uploaded to the resource archive or a non-digital resource has been ordered, you or the Source Acquisition Specialist should change the status to “Completed”. This will instruct the PanLex Bot to register the source in the PanLex database. The bot will report the outcome in a comment on the subtask after about 8 minutes. If registration was unsuccessful, the comment will contain an error message and the status will change to “Error”. Once you have fixed the problem, change the status back to “Completed” to try again.

Amend a completed source

If you discover any changes that need to be made to a source once it has already been registered (for example, after the ordered resource arrives), change its status back to “In Progress” and make the changes. Then change its status to “Completed” to update the registration in the database.

Note that you must always wait a few seconds after changing the status to “In Progress” before changing it back to “Completed”, to ensure that the PanLex Bot notices the change. When the bot has noticed, it will post a comment saying “Previously created source being worked on: changes will update existing PanLex source”. It is then safe to change the status to “Completed”.

Complete a language

Once you reach the point of diminishing returns in your search for resources for a language, it may be time to move on to another language. You can indicate that an adequate number of sources have been registered by changing the language task status to “Completed”.

Give up on a language or source

If, after appropriately exhaustive searching, you are unable to find any sources for a language, you can indicate this by changing the language task status to “Dead End”.

If a resource turns out to be inappropriate for some reason (for example, there is no way to acquire it or it does not contain the expected language), you can indicate this by changing the source subtask status to “Dead End”. This is preferable to deleting the subtask, since deletion would leave no record that the resource had been considered.

Create a new language or country

If you would like to add a language that does not exist in our list, you can do so by creating a new task under the relevant country. If the language is spoken in more than one country, you can add the other countries in the task detail view.

Our convention for language task titles is “Name [aaa]”, where “Name” is the name of the language (generally following the Glottolog name) and “aaa” is its ISO 639-3 code. If you do not specify an ISO 639-3 code in the task title, the PanLex Bot will not be able to find the task for the purpose of automatically associating source subtasks (see above).

If the only available ISO 639-3 code is significantly broader than the intended language, you should instead use the PanLex UID in the task title (“Name [aaa-000]”). For example, if the language task is for a particular variety of English rather than English in general, you should not end the task title with “[eng]”. Otherwise, all source tasks with English in their language variety list will automatically become subtasks of that language task. This is undesirable, as it clutters the list of subtasks with irrelevant sources.

If one of the countries in which the language is spoken does not yet exist in our list, you can add the country as a subfolder under “Acquisition Languages”. To do so, right-click “Acquisition Languages” in the far left menu and choose “Add > Folder”.