Carlos García Montoro
I'm a newbie Refbase world. My institute wants to manage the literature that its members produce and I have proposed Refbase as a way to do it. One of the problems that we have is that we want to categorize the articles in different (public) groups, according to the different research areas of our institute. I have read some work in progress of other people with this goal. Since the current groups seem to be personal groups, I believe that they don't really help to categorize our publications.
But there is a thing that at the moment I'm not able to see its current utility. If you login refbase with the administrative account, you can "Manage Users", and there you can select some users and add them to groups. When I saw this feature, I though that I could do some groups to share group-based literature. Users are members of groups and the can see / add / remove articles to their groups. Then I noticed that you cannot see group membership and finally I saw that you cannot see / add / remove articles to user groups. I haven't found a single page other than "Manage Users" which uses the user groups. So my question is, for what are they intended to?
Any advice is always welcome,
Do I understand you correctly, that you want to have the articles in those groups only visible to designated members?
If not, i.e. if all articles in the refbase database are visible to all users, then I dont really see the problem, since all articles can be sorted in any way you like in refbase. In other words, every member of the workgroup "Drosophila" can make his/her own private group called "Drosophila" - still everyone is being notified about the same articles that are being uploaded…
Not really. Articles can be public, but the have to be classified according to our institutional groups. I don't care if any user makes it own categories (i.e. groups) for his/her own purposes, but I wanted some pre-established (global) groups for all the users to classify articles. I thought groups could help me, but they are different for every user. Then I thought "user groups" was the solution, but I'm not able to see how they are used in refbase.
On the other hand, I have seen that all this info is present in the database. So perhaps, some customization could be done…
Thank you for your answer. Hope my question is more understandable right now.
hmm..OK, I see your point. Also, I don't understand the use of the User-groups feature in RefBase.
One way to organize users and especially the publications of users would be in the user settings. You have the options: Surname, Name, Institution and Abbrev. Institution. If each user for the institution drosophila has the institution "drosophila" in the user settings, maybe that could help. articles being uploaded could be marked to belong to the institution "drosophila" for example, and other articles belong to institution "genetics" and so on…
OK, I just tried it out: I added each user to an institution, p.ex.: Institution: Workgroup Drosophila; Head Institution: INstitute for Neurobiology …
When a user now uploads a reference, there is a small checkbox at the end of the form:
(in this case, the User belongs to "AG Stein".
I think that would be a way to sort your references…
the other question is, if you can actually use it for searching and stuff, but I think there is a how-to in the refbase docs which explains how to create user/group references….
The groups found under "manage users" is to handle a group of users. Presently, it is mostly useful only for administrators who have to handle a large number of users in different groups. More functionality involving groups of users (rights management, etc.) are planned for the future. This is all likely a red herring to what you'd like, though.
You want to handle groups of references, not users. you are correct that "user groups" on the various pages dealing with references is peculiar to the single user who is logged in. For small workgroups, some people have created a single refbase account in order to share this information across all users. If you need separate user accounts, this is not an option. Having this kind of group/tag shared across users is planned and a bountied developer is apparently going to work on it. There are various threads here on that topic. The work-around is to use some field for this that is already public. "Keywords" is a relatively good one to use.
Carlos García Montoro
Thank you for your answer,
florianneuro, some people of my institute pertains to more than one research line (more than one group). I made use of the field that you mention in order to recognise the publications of our institute, but I cannot put there the group information since some researchers have more than one group. A you stated, using this field, you can "Library Search" the internal publications of the institute.
karnesky, thank you for your comment. Your idea is good. With a single user you overcome this kind of problems. My institute has about 200 members. I don't think that my boss will be happy with this solution, but I'll mention it.
There is a way to do it. If I populete the group combo box of each user with my groups, they can check articles as being related to their groups. Nobody else that the user who checked it is able to see this, but if a user has administrative rights, then he can perform the following sql search:
SELECT author, title, year, publication, volume, pages, thesis, orig_record, serial, file, url, doi, isbn, type, orig_record, serial, file, url, doi, isbn, type FROM refs, user_data WHERE refs.serial = user_data.record_id AND user_data.user_groups LIKE '%Group 1%' ORDER BY author, year DESC, publication
Using this query I can get every publication tagged as "Group 1", ignoring the user who tagged it. It isn't the finest way to do thing, but at least I can collect this info.
Now I have to learn how to populate this combobox.
If I were able to bypass the "user_id" when recording and/or retrieving the "user_groups" in the "user_data" table, then I would get the same results.
Any further help is welcome.
Sorru for the mess. The SQL search statement should be: