Feature of the week

Today I want to start a series of blog posts which explain features of BibSonomy which may be not so well known or understood but nevertheless be helpful for our users.

I'll start with a feature we added last week and which enhances the possibilities when uploading BibTeX files to BibSonomy. Often your BibTeX entries already contain tags inside keywords or tags fields. BibSonomy then adds these tags to your entries. Unfortunately, sometimes the tags in these fields are not separated by whitespace but by comma or semicolon. Hence we implemented the option to use another delimiter than whitespace as tag separator when uploading a BibTeX file.

You can find this option below the file upload form on the post bibtex page. Next to the upload button you can find now an options link. When you click on it the following options appear:

• character encoding
• viewable for
• tag delimiter

with the latter being the one I'll explain here. There are essentially three things you can change for this option:

• The checkbox allows you to enable or disable this feature. By default this option is disabled, which means that tags contained in the BibTeX fields tags or keywords are assumed to be separated by white space. If you enable this checkbox, you can configure with the dropdown box the character which separates the tags.
  
• The dropdown box allows you to choose between comma and semicolon as separator. Picking the comma, for example, means that the string contained in the keywords or tags fields of the file's BibTeX entries will be cut into several strings at each occurrence of a comma. This results in several tags which still may contain whitespace because the string "computer algebra, math" would be cut into "computer algebra" and "math". Since BibSonomy does not allow whitespace within tags we have to take care of the space in "computer algebra". Leaving it as is would result in the three tags "computer", "algebra" and "math" -- which may be appropriate or not, depending on the users choice. But the next options allows us to control this behaviour.
  
• The last input box allows to specify a character which will be used to join several words of one tag into a one word tag. The default is "_" resulting in "computer_algebra" for the preceding example. Removing this character completely would result in "computeralgebra" and entering a space would give us two (!) tags "computer" and "algebra". Hence this little input box gives you some power regarding what your tags will look like.
  

I want to finish with the remark that this option may look a little bit complicated which results from the power it has. Therefore I propose to use it first only on small examples to understand how it works. If you have questions or comments, don't hesitate to comment in this blog or contact us.

Robert

Feature of the week: interactive tutorials and new help structure

On 12th April 2007 it is BibSonomy's birthday :-) Two years of design, development and the contribution of its users have enhanced the system constantly. In order to show BibSonomy's features and support users starting to work with the system, we have changed our help-websites.

* A general overview of all help functionality we offer can be found at:
--> http://www.bibsonomy.org/help/

* A special highlight are the interactive tutorials making basic and complex actions in BibSonomy easy to understand (thank you Christian!):
--> http://www.bibsonomy.org/help/tutorials

* Users who want to get the most of BibSonomy can find tips, tricks and useful tools at:
--> http://www.bibsonomy.org/help/addons

* Finally, we want to give more background information on what is going on behind and around BibSonomy:
--> http://www.bibsonomy.org/help/about

With the new help features, we hope to improve BibSonomy's usability. We are happy to get your feedback if important information is missing or if you have more ideas ...

Beate

Feature of the Week: Customizable Publication Exports

Many of you may already be using the export facilities for publication entries that BibSonomy offers; until recently, there were options to export your BibTeX as RSS, RDF, Endnote, or one particular plain HTML format.

Now we've revamped the export facilities by including the export layout engine from the great JabRef stand-alone BibTeX manager. Clicking on the more button next to the publication column header will take you to a list of publication output formats that are now available—right now there are 17! These include an RTF export for standard text processors such as Word, CSV which is easily imported to spreadsheets, as well as fancy HTML tables with sorting capabilities that you can integrate right into you web site.

Here are some examples:

• CSV for spreadsheets
• Fancy HTML table
  
• RTF for Word and other text processors
  
• DocBook XML

If there is still a layout missing that you need for your particular application, you can make your own! The JabRef documentation describes how custom export filters can be written using a special markup language. Once you've written your filter, you can upload three files: the piece of markup for formatting each entry, plus a custom header and footer; thus, you can get a complete, valid HTML page with your custom publication list out of BibSonomy which can directly be linked to without further processing. The URL scheme for these layouts is as follows: http://www.bibsonomy.org/layout/[filter]/[rest of BibSonomy URL] e.g. if I want to format my own publications of 2006 with the html filter: http://www.bibsonomy.org/layout/html/user/schmitz/myown+2006 See also Mark Schenk's page for some great examples of JabRef layouts.

Have fun!

Christoph

Feature of the Week: Groups and Friends

If you have bookmarks and publications, that you want to share with specific people only, you can define them as being your friend (on the friends page). Then you can post selected bookmarks and publications to your friends only. On your friends page, you can see whom you have declared being a friend up to now. You will also see if others have declared you being their friend.

Groups extend this idea of collecting and/or sharing resources. There are two aspects of groups. First a group can be used for aggregating the entries of a specific group. An example is http://www.bibsonomy.org/group/kde/myown which collects all entries that are tagged with "myown" by at least one person of our research group, and which we use for generating our publication page http://www.kde.cs.uni-kassel.de/pub. Second you can restrict access to posts to your group. If you select a group when posting, only members of that particular group will be able to see that content.

The posts still belong to the original user. This means that if the user leaves the group, or cancels her account, the information gets lost for the group. In order to prevent the loss of group knowledge, posts can be automatically copied to one or more groups which the posting user is a member of by attaching the special tag "for:username" to the post. This causes the post to be copied automatically to the respective group, with an additional tag "from:username". This function is also useful if you need a mechanism for finally committing entries to someone else, eg as project deliverable, see http://www.bibsonomy.org/user/nepomuk (which is reused here) for an example.

For turning a normal user account into a group account, write an e-mail to webmaster@bibsonomy.org . We will then make this user the group admin.

Feature of the Week: Information Extraction supports the Import of References from Homepages

Todays feature of the week post will point you to one of the hidden features of the system. As most of you certainly know one way to acquire the meta data of a publication is to use the screen scraping facility of BibSonomy. A list of supported sites can be found here and is extended constantly. Today we released a new scraper for Highwire and LibraryThing. It's also possible to write your own extension. A description of the internal scraper interface is provided here and allows you to implement scrapers for BibSonomy.

At the end of the list you find the IEScraper which is not designed for a special web page but rather supports you in general by the import of "usual" formated publication metadata like the following one:

Emma Tonkin and Marieke Guy. Folksonomies: Tidying Up Tags? . D-Lib,volume 12(1), January 2006.

which you can find at:

http://www.cs.bris.ac.uk/Publications/pub_info.jsp?id=2000478

To use this scraper you have to highlight the text of the reference you like to copy and then press the post_publication button. What happens in the backgroud is: The marked reference is send to the BibSonomy server and as no other scraper is able to process this kind of entry the IEScraper processes the entry and tries to find the different parts of the reference like: author, title, or year. You end up in the publication input mask where you find a prefilled form containing all information the scraper was able to extract. Now you can add your tags and adapt the entry. As an example the above entry in BibSonomy:

http://www.bibsonomy.org/bibtex/29488117bf156fe15b2fb3b8ab4376dec/hotho

Unfortunately the information extraction technology is not able to process all entries correctly. For the following entry:

Philipp Cimiano, Andreas Hotho, Steffen Staab. Comparing conceptual, partitional and agglomerative clustering for learning taxonomies from text. Proceedings of the European Conference on Artificial Intelligence (ECAI'04). 2004.

title and authors are extracted correctly but the booktitle is wrong. It contains the missing year, too. You have to correct this mistake manually. We are logging this correction and using this kind of information to tune the IEScraper. Currently we have to start the training process manually but we are working on an automatic learning setup.

We hope that this feature supports everybody who finds references not at the common digital archives but rather at homepages of researchers. As the IEScraper is not perfect it takes over a reasonable amount of the work and we hope you find this feature useful.


Have fun!

     Andreas

Feature of the Week: Relations

Sometimes I want to have some means of structuring my tag cloud. For example, I tag a lot of researchers' home pages with their last names as tag. So I want to have all these last names together in one spot.

Furthermore, why do I have to say event every time when I tag something with conference, just so I can also retrieve it under the more general term?

Relations

To solve both problems, BibSonomy has introduced relations. A relation between two tags signifies that one of the tags is regarded as a generalization of the other.

When I tag something with conference, and I have the relation conference->event stored in BibSonomy, I say that I also consider that resource I tagged to be concerned with event in some sense.

Relations can be entered just in the way shown above: just write an arrow "->" or "<-" between two tags. The tag pointed to will be the more general one. This can also be chained: you can tag, say, a page related to pet cats with cats->pets->animal.

We call those tags that are more general than another tag, i.e., those at the pointy end of an arrow, concepts.

Using Relations

Now how can relations be used? First, they appear in your tag box on the right hand side. Note the upward and downward arrows next to all tags that are concepts. Clicking on an upward arrow shows the relations of that concept, clicking the downward arrow hides them.

Second, you can use concepts to retrieve posts. Note that when on a tag page, such as http://www.bibsonomy.org/user/schmitz/linux, you will be offered the possibility of seeing the tag linux as a concept. That way, you will also see those resources not tagged with linux themselves, but with a direct subtag thereof, for example ubuntu (assuming I have the relation ubuntu->linux in my relations).

Editing Relations

Entering relations can be done on the fly, with the arrow notation shown above. They can be browsed under the myRelations link. If you want to edit your relations, hit the edit tags link in the top right corner. You will be offered the option of adding and removing relations. Furthermore, note that when renaming a tag (say, you've made a typo and spelled ubutnu instead of ubuntu), there is an option on the edit tags page to update relations containing that tag, too.



Have fun,

Christoph

Feature of the Week: Keep and collect interesting publications in your basket

For collecting publications from other users, BibSonomy offers a basket function. To add entries to your basket, a pick function has been attached to each BibTeX entry. If you are interested in a specific BibTeX entry, you can choose the pick option and the entries are added to your basket.

Your basket can be displayed with the basket button at the right menu. On the following page, you can select or remove items from the basket, and export the final list as BibTeX and EndNote format. Additionally, you can update the tags of your BibTeX entries in a batch mode. This option disregards BibTeX entries, which are not your own to avoid changing tags from foreign BibTeX entries.

Try our basket option and have fun, Miranda!

Feature of the week: Firefox integration

BibSonomy offers several facilities to retrieve bookmarks and publications from personal or public collections. An example is the fulltext search mentioned earlier in this blog. For everyday work, it is of great use when these facilities are integrated tightly into the usual working environment, e.g. into the accustomed web browser.

As most of BibSonomy's retrieval features can be accessed via a specific URL structure, the built-in keyword replacement mechanism of Mozilla Firefox provides an elegant mechanism for this purpose. To give an example, if you want to search bookmarks or publications inside your own collection that were tagged with a particular tag, you would have to do the following:

1. Create a new bookmark inside Firefox (via the usual Bookmarks > Bookmark Manager facility)
2. Enter the following details:
   

• Name: Search My BibSonomy
• URL: http://www.bibsonomy.org/user/[your_username]/%s
• Keyword: bs
  

Once this is accomplished, the personal collection can be searched directly via the Firefox address bar: Jump to the bar (e.g. by pressing CTRL + L), and type "bs" (i.e. the keyword that you just specified when creating the bookmark), followed by a space-separated list of the tags you would like to search for, e.g.

bs myown 2006

to retrieve all bookmarks and publications tagged with "myown" and "2006".

This method is obviously rather flexible and can be applied analogously to integrate e.g. global fulltext search. For this purpose, the URL in step (2) from above would have to be specified as http://www.bibsonomy.org/search/%s.

In combination with the keyboard shortcut to post a bookmark, a tight integration into Firefox is accomplished that eases every-day publication and bookmark management with BibSonomy.

Feature of the Week: Author Search

One of the recently released new features is the author search. The author search is implemented based on the MySQL full text search feature of the myisam database engine. What we are doing is that we copy all the author information of a publication into a text field of a myisam table. Based on that we are able to request author names as words. This is a very fast and simple way to implement this feature.

The simplest way to search for an author is to try to search for the last name like Knuth. That's what we support right now and what we call a author page. What you get then is a list with all publications and a tag cloud of the author describing the topics of the author based on the tagged publication of our users. Every publication item contains now a link on the last name of every author linking its author page.

An additional restriction of the author search is possible by adding an additional author like  Janson or a tag restriction, see http://www.bibsonomy.org/author/Knuth/MMIX.

Currently we do not support a restriction by the first name of an author. An author name disambiguation is also not possible as we do not have a separate author table where we could store the same name for different author. We are planing to extend the current author search in those directions by a more advance version in the near future.

Have fun
  Andreas

Feature of the Week: Automatic Detection of Scrapeable Content

Sometimes when you surf ACM, Springer or similar sites, you inadvertently punch the “postBookmark” instead of the “postPublication” button when you really want to post a publication.

BibSonomy now automatically detects if you are on a site it has a screen scraper for, and offers the possibility to choose whether you want a bookmark or publication post.

myBibSonomy-Menu: Manage your Personal BibSonomy Data

After last week's short description of the new BibSonomy release's changes, this entry will explore one feature in detail: the menu under "myBibSonomy". Basically, this menu offers different perspectives to view and manage your personal BibSonomy data:

* Yep, clicking on the first link (myBibSonomy), you still get to see your bookmarks, publications and the tags you have used so far.

* myFriends allows to share entries with people you trust.

* myRelation presents all relations you have defined so far. You can add more, edit or delete relations as you like.

* The mySearch feature offers a fast search in your collection. You can specify tags and authors from a list containing your personal tags and authors. When you chose the "and" option, you can mark several tags and authors by pressing the CTRL button. The system will return all entries which can be found with the given tags and which were written by the specified authors. If you still want to reduce the results, you can filter the entries by entering the title's keywords, tags or authors in the "filter" textfield.

* myPDF forwards you to an overview of those personal publications to which you have attached a document (pdf, ps or DJVU).

* Finally, myDuplicates shows duplicate entries within your publication list.

Beate

Feature of the Week: Manage the publication lists of your homepage, research group homepage and research project homepage with BibSonomy!

Researchers usually have to report the publications they have written several times, often in different formats. Did you know that BibSonomy simplifies this tedious work? Stop reformatting and restructuring the list of your publications again and again for all the different web pages you usually have to maintain - e.g., your personal homepage, the homepage of your research group and/or department, the homepage of your research project, ...

How to do it? See some examples:
1. Personal tag clouds and publication lists. The tag cloud on my personal homepage and my publication list are generated from BibSonomy. The key idea is to use a specific tag in BibSonomy to mark all publications one has co-authored - in our case the tag "myown".
The web server of our research group queries every half an hour the URL http://www.bibsonomy.org/user/stumme to update the tag cloud and http://www.bibsonomy.org/publ/user/stumme/myown+2007?items=1000 (and similar pages for the earlier years) to update the publication list. The relevant HTML part is extracted, e.g., for the tag cloud everything between the start pattern <ul class="tagcloud" id="tagbox"> and the end pattern </ul>. The extraction process depends on the functionality of your web server, hence we cannot give general hints here. In our case, the formatting of the final tag cloud/publication list is realized with a css style sheet which is placed on the server of our research group. Instead of doing the layout of the extracted HTML on your own web server (eg by using css), you can alternatively export it directly in the desired format from BibSonomy, see "Feature of the Week: Customizable Publication Exports" for more details.

2. Publication lists of research groups and departments. The same data are used for displaying the publication list of our whole research group. This list is extracted from http://www.bibsonomy.org/publ/group/kde/myown+2007?items=1000 by selecting everything between the start pattern <h1>Publications</h1> and the end pattern </body>, as described above. In general, the group functionality allows to aggregate all entries (or only those entries tagged with a tag like "myown") of a group, see "Feature of the Week: Groups and Friends" for more details.

3. Publication lists of research projects. You can again reuse the same BibSonomy data for generating the publication lists of research projects. Two examples are the publication lists of the European projects TAGora and Nepomuk. For doing this, the project coordinator sets up a group account in BibSonomy, and the project partners should join this group (see again "Feature of the Week: Groups and Friends" for more details). Then there are two alternatives, depending on the requirements of the project: a) Define a specific tag, e.g. "tagorapub" as done for the publication list of the TAGora project. An author devoting a publication to the project marks the publication with the specified tag, and the project web server collects the data from BibSonomy (in the format that is most suitable for your web server, eg. from http://www.bibsonomy.org/swrc/group/tagora/tagorapub). b) Ask each project partner to tag the publication he devotes to the project with "for:". An example is the publication list of the Nepomuk project. The advantage of the second method: Once a publication is commited to a project, it cannot be withdrawn without explicit consent of the project coordinator. This feature is often required in projects where the coordinator is responsible for the reporting.

Feature of the Week: Tag Box Settings

Some weeks ago we improved and unified the settings for the tag cloud. If you have JavaScript enabled you see now above your tag cloud the options for sorting (Which could be either by frequency (i.e., how often you used a tag) or alphabetically.), layout (You can choose between a cloud or a list.), and minimal frequency (Allows you to show only tags which you used at least that often.).

On the settings page all those options (and some more) can be changed, too. There you can also adjust the minimal tag frequency at a finer granularity and activate tooltips for your tags. For users with deactivated JavaScript the settings page is the preferred way to change the layout of their tag cloud.

Furthermore, if you are logged in the settings are now saved in the database. Hence, when you switch to another browser or computer, you see your tag cloud always in the same layout.

Feature of the Week

Extend your publication entries by adding a .pdf/ps file and a private note

Besides assigning tags to your BibTex entry with BibSonomy, you can extend your entries by uploading a local copy of publication as .pdf or .ps document. In addition, you can assign some private notes, which are only visible for you. This notes can contain some personal thoughts like “this paper is important for my research” or a short summary about the content. For example: You just posted the following publication: “Network Properties of Folksonomies” and would like to extend your entry with a local copy as .pdf or .ps file. You just simply click on the title and get the following options (see figure at the top). Try it and have fun, Miranda!

Feature of the Week: New URL schemes

We have implemented some new URL schemes to make your life easier:

Reference by BibTeX key. If you are looking for a particular entry for which you know already who has stored it (eg, yourself) and for which you know the BibTeX key, you can directly refer to it with the following scheme: http://www.bibsonomy.org/bibtexkey/KEY/USERNAME . For instance, check out http://www.bibsonomy.org/bibtexkey/Wille82/stumme. If the user has more than one entry with the same BibTeX key, then a list of all hits will be given. You can use this feature also for referring to entries that we mirror from DBLP, once you know how DBLP generates its BibTeX keys, eg, http://www.bibsonomy.org/bibtexkey/journals/jacm/HopcroftU69/dblp.

Short-Cuts. The following links are available only when you are logged in. They do not provide new functionality, but are convenient short-cuts.

• direct link to the list of your own bookmarks and bibliographic references: http://www.bibsonomy.org/myBibSonomy
• direct link to the BibTeX list of all your own bibliographic references: http://www.bibsonomy.org/myBibTeX
• direct link to the list of your own relations: http://www.bibsonomy.org/myRelations
• direct link to your own collection of full texts (PDF, PS or DJVU): http://www.bibsonomy.org/myPDF
• direct link to your own duplicates: http://www.bibsonomy.org/myDuplicates

Feature of the Week: Fulltext Search within Tags:

Our fulltext search contains an addition helpful feature to support the user to navigate through the content of BibSonomy. The tradition fulltext search finds words contained in URLs, titles, descriptions and especially all BibTeX fields like author, editor or bibtexkey. Now, we improve the fulltext search functionality to find also words, which match with tags (see figure at top). Use it and find more bookmarks and publications in which you are interested. Miranda

Feature of the Week: Content negotiation

Content negotiation helps to represent a resource at the same URL in different ways considering the capabilities of the requesting user agent. It allows user agents to choose between several representations of a resource by giving an appropriate MIME type in the HTTP-Accept-header.

Browsers typically accept MIME types of text/html to get some human readable representation. A semantic web application, though, might prefer to get a representation of the resource in RDF and thus accepts only application/rdf+xml. In BibSonomy we have added this behaviour to URLs representing particular resources - i.e., a single bookmark or publication reference (which may be represented by several posts). You can access this feature by prepending the /url/ and /bibtex/ URLs with the prefix "/uri/", e.g.,

http://www.bibsonomy.org/uri/url/d1bb7b3f6cafafa7b418f9f356ff2e83

or

http://www.bibsonomy.org/uri/bibtex/2b8b87c78e9e27a44aacde0402c642bff

Depending on the Accept-header of your user agent you get a redirect to the appropriate representation of those resources. You can read some more information on this in the BibSonomy help system.

Feature of the Week: API and JabRef Connectivity

People have been asking us about this for a while, and now we are happy to announce it: BibSonomy now has an application programming interface (API) which allows you to code external applications that interact with BibSonomy.

So for example, assume you have a standalone tool for managing BibTeX entries which you use on a daily basis. Now you want your tool to be able to interact with BibSonomy, so that you can publish references from your desktop to BibSonomy at the push of a button.

JabRef: Standalone BibTeX Manager now Interacts with BibSonomy

Actually, this is what we have implemented as a first application ;-) Have a look at the API download page. There you will find a customized version of the great JabRef reference manager which has been extended to interface with the BibSonomy API. This version is able to publish your entries to BibSonomy, retrieve references with a specific tag, or modify references in a fully-fledged desktop application rather than using a web interface.

This also means that you can use your references on your laptop, say, when you do not have network connectivity, and synchronize them with BibSonomy when you reconnect.

REST API for Programmers

For the more technically minded: the API offers a REST-type interface to BibSonomy resources. So for example, to get all users, you'd say "GET /users" over HTTP. To modify a particular user, you'd say "PUT /users/<username>" with an appropriately formatted XML document with the user data. The API documentation documents the possible operations and their current status — not all of them are finished yet.

So this allows you to code an application against BibSonomy in about any programming language, although you'd have to write all the HTTP and XML wrangling yourself (at the moment). For the Java language, we are also offering a client library which lets you write applications against BibSonomy that handle elements of BibSonomy, such as posts etc., as proper Java objects and hides all the gory details.

Getting an API Key

Of course, all this only works within the bounds of proper authorization. To be able to use the API, you have to obtain an API key: just check the second settings pages.

Feature of the Week

As explained in the last feature of the week, BibSonomy allows users to structure the content via SUPERTAG <- SUBTAG relations. The built tag concepts are available for searching and navigation through our folksonomy system.  As seen in the figure above it is ease in handling, whereas each step is symbolized with a circle. Only choose “concepts” (step 1) as search option and type a tag which your are interested in (step 2). In the last step (step 3), you get resources and a visual presentation of your concept as hierarchy.  Miranda

Feature of the week: Retrieve resources by disjunction of tags

A very common way to browse through your own or other people's repository on BibSonomy is via one or more tags, e.g.

http://www.bibsonomy.org/tag/semantic+web.

Hereby, the tag-based retrieval is done in a conjunctive manner, i.e., the result of this query will comprise all bookmarks and publications tagged with semantic AND web. We are often asked if we offer any other possibiliby of combined tag queries, e.g. by disjunction - one might e.g. be interested in all resources tagged with semanticweb OR ontologies.

This behaviour is not accessible via a specific URL scheme, but can be achieved by invoking an old BibSonomy buddy - namely concepts! As you will know, BibSonomy allows you to define relations between tags in the form:

SUPERTAG <- SUBTAG

(see also http://www.bibsonomy.org/relations). A supertag along with all its subtags is denoted a concept in BibSonomy, which can be used to retrieve resources like this:

http://www.bibsonomy.org/concept/tag/ontology

The characteristics of this retrieval method is now that all resources are returned which are tagged with ontology OR one of its subtags. This constitutes, in fact, a retrieval of resources by a disjunction of tags. We are aware that this has some limitations, as a concept has to be defined before this type of query is possible - but facing a tradeoff between efficient query processing and freedom of query formulation has led us to this decision, with the ultimate goal to keep our service highly responsive for all of you as our users.