Feature of the week: Bibsonomy's new Layout

As you probably already noticed: BibSonomy got a new outfit. We will summarize how look & feel of your (still) blue social bookmark and publication sharing system has changed.

Navigation
Little changed with respect to BibSonomy's menu navigation, though, due to extensive use of cascading style sheets (CSS), overall navigation is more consistent and obvious in comparison with the old layout.

Whenever a menu item is linked to a sub menu, a downward triangle is displayed left to the corresponding item. The main menu realizes tab-like navigation: While visiting a certain page, the corresponding menu item is displayed in the front.


Tag clouds
The picture above also shows the new tag cloud representation: The higher a tag's frequency, the bigger its font size and more intense its color.


List views
Each list view which shows publications as well as bookmarks offers the possibility to expand one of it's columns. This feature can be found on the top right:

Double column view is restored by activating the '|' link at the same place.

Quick-Post
You can tag any post displayed by your own by clicking on the star on the right of the corresponding entry in a list view. A solid blue star indicates that you already tagged the resource.


Internationalization
Internationalization of BibSonomy is still in progress. Whenever available, different languages can be chosen by clicking on the corresponding flag at the top right corner.



...and all the rest
Not each element's new layout has been introduced here (e.g. the new groups listing, just to mention one). Have a look by yourself and enjoy BibSonomy's new layout.

Errors tend to happen, though we try our best to avoid them. If you encounter any incorrectly rendered elements, please let us now by sending an email (see our contact page for further details).


.folke

Feature of the week: OpenID support in BibSonomy

OpenID is an open, decentralized standard, which allows users to log onto many different services on the web using the same identity identification (single sign-on). This kind of authentication is provided by a growing number of websites, including large ones like AOL, Google, Microsoft, MySpace, Yahoo and many others. And now for the good news - we are happy to announce that BibSonomy is among those!

OpenID makes registering at BibSonomy a lot easier. If you already haven an OpenID, you can go directly to our register page and click the link "register using OpenID" at the bottom:
In the next step, you enter your OpenID:
Then, you will have to confirm your BibSonomy registration at your OpenID-Provider (in this example myopenid). When you trust BibSonomy, you should activate the checkbox "Skip this step next time I sign in to www.bibsonomy.org" - then you won't be redirected to your OpenID provider the next times you log in to BibSonomy:

And finally you can complete your user settings. Please note that your name, gender and other information is fetched from the OpenID-server, so you don't have to retype all those details each time your register:
The next time you want to log in to BibSonomy, you can do this by clicking the link "Login with OpenID" at the bottom of our login page.

If you do not know your OpenID, it might nevertheless be the case that you already have one, e.g. when you have a Flickr account - refer to the official OpenID page to find out more details on this. Please note that OpenID will On the same site you will find information how to obtain a brand new OpenID, which can then then make your online experience easier and better - including BibSonomy!

Best,
Dominik

Metadata Scraping Service

As mentioned in Robert's last blog post we set up a scraping service which supports users working with citations by extracting automatically references from digital library or publisher websites. We use a very similar service in BibSonomy to support our users while posting a new reference. However, the service is independent from BibSonomy. Our main goal is to make the metadata of other websites easily accessible to every user who needs bibliographic metadata. Therefore we offer the extracted information in BibTeX format. Most tools allow to import BibTeX so it should be very easy for everyone to get the data into his own tool. The service is running under the following URL:

http://scraper.bibsonomy.org/

Currently we support more than 60 different websites (here the full list) and we are working on further extensions. In the near future we will make the source code of our scrapers publicly available under GPL and we hope that other people will find it useful and start to help us by implementing their own scrapers.

How does the service work?

In principle there are two ways to use the service. One uses a so
called bookmarklet and the other is simply based on the URL. If you
have a webpage of a supported site e.g. from ACM digital library the
following page:

Logsonomy - social information retrieval with logdata

then you can copy this URL into the form on the service homepage and the service will return you the extracted BibTeX information. As this is not a very convenient way to access the data we provide a ScrapePublication button. This button is a small piece of JavaScript and can be copied to the toolbar of the browser. By pressing this button while visiting a digital library webpage the URL will be automatically copied and sent to the scraping service and the metadata is extracted.

The service has three options which can be used to customize it and to make it useful for other systems. Obviously one parameter is the URL itself which is used by the bookmarklet, too. The next is the selection parameter which allows to send text to the service and the last parameter allows to change the output format from html to plain BibTeX. This last parameter makes integration with other systems very simple.

If needed we can provide the metadata in other formats as well but currently we support only BibTeX.

Feature of the week: clouds

Looking out of the window, I can see the visible mass of condensed droplets or frozen crystals that are so typical for a German sky in autumn. Consistent to the weather, this week's feature is about clouds. In BibSonomy (and other collaborative tagging systems), clouds reflect the common vocabulary of the system's user. It's fun to occasionally viewing BibSonomy's cloud to get an overview of what other people liked. When hitting the cloud link, you may have realized some layout changes. Also, the computation of the tag cloud changed. We now exclude predefined tags such as imported, and only consider those tags that have been used by more than 10 different users. Small changes, which hopefully make a difference.

Beate

Feature of the week: status update

Sorry for the sparseness of our "features of the week", we were traveling, on holiday and implementing new features. We will describe new features in the next weeks in more detail. I will give you a short overview today:

• Links to documents attached to publication posts are now included in the API's XML (including the MD5 hash of the file!).
• BibSonomy now supports OpenID.
• A new popular page and an updated tag page.
• Logging of click events in the web interface. This is for research purposes, e.g. to better evaluate tag recommenders. You can disable this on the settings page.
• We're currently working on an open tag recommendation interface to allow interested researchers to evaluate their recommenders right inside BibSonomy.
• We will shortly make the code of our scrapers publicly available and since today provide a simple web service which allows to use the scrapers independent of BibSonomy.
• A new version of the Java client library is available. We also published the corresponding JavaDoc.
• And last but not least: numerous bugfixes went into the new releases.

So we're looking forward for the next weeks to explain new or updated features and also to new features we will release shortly. There are some interesting things in the pipeline!

Feature of the week: GoogleSonomy - A Firefox addon to enhance your BibSonomy Search

GoogleSonomy is the name of our new Firefox extension. As you can see in the screenshot below it shows you BibSonomy search results along your Google search. You can retrieve four different lists of BibSonomy posts for a google query. Those are tag and fulltext search, each on a global or a user base.



In the settings dialog you can specify which lists you want to retrieve and how many entries should be displayed each time. Furthermore you can set your BibSonomy user name and choose a default search mode.



To obtain this extension visit Mozillas Addon Page. Unfortunately you need to be a registered user to install GoogleSonomy because it is like any other new firefox extension in an experimental state. Only with some reviews from users we can achieve our aim to get this extension in the public download area of Mozillas download page. So we would appreciate your effort.

If you want to save pains just use this link to install GoogleSonomy in your Firefox web browser.

If you have suggestions how to improve GoogleSonomy or if you found any bugs, please let us know.

Best,
Stefan

Feature of the week: Updated JSON feed

In one of our last posts, we announced the availability of a JSON feed in BibSonomy. It was meant as a lightweight way to get tag clouds for publications only. Since popularity of JSON increased and we got requests to include feeds for other pages as well, we decided to implement JSON feeds for all pages. Those feeds are now available!

For every BibSonomy page you can get a JSON feed of it by prepending json/ to the path part of the URL, e.g., to get the JSON feed for http://www.bibsonomy.org/tag/json, use the page http://www.bibsonomy.org/json/tag/json.
This returns you an Exhibit compatible JSON feed which includes all bookmark and publication posts of the respective page. To include the JSON feed into your Exhibit, add a link to it into the header of your Exhibit HTML code:

<link href="http://www.bibsonomy.org/tag/json?callback=cb" 
type="application/jsonp" 
rel="exhibit/data" 
ex:jsonp-callback="cb" 
/>

Have a look at the list of Andreas' publications to see, what is possible using JSON and Exhibit with only some lines of HTML!

Users of our existing JSON feed should note that the format changed slightly to adopt it to Exhibit's conventions:

• The tags are now contained in the list with key items and not tags.
• The name of the tag is expressed using the key label and not name.
• There are two more lists (types and properties) to allow easier Exhibit integration.

We've already updated the available tag cloud script to fit to the new layout.

Feature of the Week: Integration into your WordPress blog

Bloggers who are using WordPress may be interested in a plugin that helps them not only to integrate data from BibSonomy into posts and ordinary pages but (almost) automatically publish their posts on BibSonomy.

Once you've installed the plugin, go to Settings - BibSonomy and at least enter your username and API key. Then test your setup with a click on Show posts, select a post and submit it with a click on Post selection - your post will appear as a new bookmark on BibSonomy.

While it's a really nice thing to have your blog posts synchronized with BibSonomy another powerful feature is to integrate your bookmarks, publications or tags into your content. Publishing your tag cloud from BibSonomy boils down to this simple shortcode: [bibsonomy-tags]. More complex requests like retrieving the three most recent publications that are tagged with folksonomy is possible too: [bibsonomy-posts resourceType="bibtex" tags="folksonomy" end="3"]. Have a look at more examples.

From a developers point of view this plugin is powered by an implementation of BibSonomy's webservice in PHP. If you want to build something on top of that, you can download the API here. We're really interested in your feedback, so don't hesitate to write us about your experience.

Making BibSonomy integration smoother: New features for the University of Cologne's library users

The idea of BibSonomy (and other social bookmarking systems) is to allow users to manage their references, but also to provide a facility to browse interesting content. This property makes social bookmarking systems interesting to digital library users. Offering a functionality to export a library's metadata into BibSonomy allows the library's users to collect metadata from their (probably different) local libraries, to describe their resources and to share the information on a wider basis than just with other local library users.

The library of the University of Cologne (KUG) is promoting the idea of integrating libraries with user-created bookmarking systems. They already offer tagging services for their local entries.
Recently, they enhanced their service by integrating the tagging information into BibSonomy and providing BibSonomy information to their users. Two major features have been implemented to realize this:

* a browser for BibSonomy sources
* a mirrow of tagging activities

An example is given with the following link:
http://kug5.ub.uni-koeln.de/portal/connector/permalink/inst006/6439/1/inst006/index.html

The main feature, storing the metadata in BibSonomy is still provided with a simple link to the BibSonomy publication metadata interface where you can add tags and cross-check the reference data fields. Additionally, the KUG page shows related tags to a resource in the right menu bar. If you click on one of those tags, you can browse further publications and bookmarks related to the clicked tag. The related resources are either retrieved from the library or from
BibSonomy. This allows users to view a broader set of resources than only the library's entries without having to switch to the BibSonomy interface. Shortly, the mirrow feature will be activated. If users then add keywords to a resource in the KUG interface, the keywords will be automatically added to their BibSonomy entries. If the title did not exist in the user's BibSonomy resource list, it is stored automatically. Thus, the user's local collection is maintained automatically in the social bookmarking system.

The integration of the library's web application with BibSonomy is based on our API.
A more detailed description of the new feature is provided here.

Thanks to the OpenBib team for developing this functionality!

Beate

Feature of the Week: Shared (PDF) document management with BibSonomy

When managing a literature database, often people want to store a full-text copy (e.g., a PDF document) of a publication in addition to its metadata (i.e., BibTex entry). To enable an easy access from everywhere to these documents, BibSonomy allows to upload a private PDF or postscript copy for each post via the post details page, e.g. for my case:

http://www.bibsonomy.org/bibtex/283b2de310e014240257ea3ee774213aa/dbenz

(Please note that only the owner of the entry can upload such a file). In the past, BibSonomy has treated these files as private only, i.e, they were only available for download to the owner himself. After several inquiries, we have opened up this restriction to groups: It is now possible to easily share PDF documents within your research colleagues. By default, when you are logged in as a group member (e.g. of the group kde), then you see a little document icon on the group page next to those entries for which any group member has uploaded a document:

http://www.bibsonomy.org/group/kde


(Please note again that you can only see the attached documents when you are a member of group kde.) You can disable this feature by appending ?showPDF=false to the URL. If you would like to restrict your view to only those documents with an attached document, use:

http://www.bibsonomy.org/group/kde?filter=myGroupPDF

Apart from that, your own documents are also accessible via your personal user page (you need to be logged in for that):

http://www.bibsonomy.org/user/dbenz
displays your own entries, along with document icons, if available

http://www.bibsonomy.org/user/dbenz?filter=myPDF
displays only those entries out of your collection which have a document attached

We think that this feature contributes to the usefulness of BibSonomy as a complete collaborative literature database tool for research groups! If you have suggestions how this can be improved, we'd be glad to hear from you.

Best,
Dominik

ECML/PKDD Discovery Challenge

Since we're organising this year's Discovery Challenge, we would like to announce the




Call for Participation

ECML/PKDD Discovery Challenge

Antwerp, Belgium, 15 Sept. 2008

This year's discovery challenge deals with two tasks in the area of social bookmarking. One task covers spam detection and the other is about tag recommendations. The dataset the challenge is based on is a snapshot of BibSonomy. More details about the tasks can be found at the challenge website.

Important dates

May 5, 2008

Tasks and datasets available online.

July 30th, 2008

Test dataset will be released (by midnight CEST).

August 1st, 2008

Result submission deadline (by midnight CEST).

August 4th, 2008

Workshop paper submission deadline.

August 8th 2008

Notification of winners, publication of results on webpage, notification of paper acceptance.

August 14th, 2008

Workshop proceedings (camera-ready) deadline.

September 15/19th, 2008

ECML/PKDD 2008 Workshop

Feature of the Week: Similiar Tags

Since this week tag pages in BibSonomy like /tag/ontology also show similiar tags. They're located below the related tags we provide on each tag page in the sidebar. You can use similiar tags for navigation when you're looking for posts related to a certain tag which might have been tagged with not the same but other, similiar tags.

More information on similiar tags can be found in the FAQ and in the online help.

Feature of the Week: Integrate BibSonomy Tag Clouds in your website

BibSonomy now offers the opportunity to integrate its tag clouds into your private website. For this purpose it provides a JSON feed containing the retrieved tags which are filled in your cloud dynamically by a JavaScript snippet. At the moment this feature is only available for the BibTeX page but will be extended to all pages offering tag clouds in the near future.

In this example you see a tag cloud with related tags to this publication.



You can choose between an alphabetical or a frequency order and define a limit of shown tags. Just follow these few steps in the tutorial to integrate such a tag cloud in your website. There you can also find a test tag cloud where you can check the functionality by specifying several arguments.

If you have questions or comments, don't hesitate to contact us.

Stefan

See the BibSonomy API in action with WizTools.org RESTClient 2.1

When developing RESTful applications or accessing such services, it can be of great help to have a handy tool to try out its different functionalities. A useful application to this end is the WizTools.org RESTClient. It provides a slim and convenient interface to test HTTP methods like POST, GET, DELETE or PUT, along with the possiblity to edit a variety of parameters for each request.

A tutorial to access our BibSonomy API with this tool is found here:

http://java.dzone.com/announcements/wiztoolsorg-restclient-21-rele

We think that this example demonstrates well both the capabilities of the RESTClient application and our API!

Best,
Dominik

Feature of the Week: BibSonomy on your iPhone

Use your iPhone for accessing BibSonomy! @BibSonomy is a website written by Nils Windisch that shows the five most recent/popular bookmarks and publications from BibSonomy, and is optimized to look good on your iPhone. Nils suggests that "the best use case is probably waiting for the bus or being bored in school seminars as @BibSonomy just gives you a quick and easy glance of what’s up at BibSonomy." Read more at http://nilswindisch.de/code/iphone/bibsonomy/.

We stronlgy encourage and support the integration of BibSonomy data in 3rd party applications, as demonstrated by @BibSonomy. If you're interested in building your own application, have a look at our API, which should provide all necessary integration facilites.

Easteregg of the Week: BibSonomy Link on Websites

Some websites such as the German news magazine Spiegel provide a link or a button on their web pages so that a page’s visitor can easily store the page in a social bookmarking system. For example, the column Zwiebelfisch about the specialities of the German language shows a "Bookmark" link above the title. Providing such a link has several advantages for providers and users: on the one hand, being visible in a social bookmarking system can increase the popularity of the website, on the other hand users get the possibility to store interesting web resources at a central place.

The technique to include such a link or button on your website is simple and can be realized by including a small javascript into your html code. For people interested in offering such a link or button to store bookmarks in BibSonomy we now provide the javascript code on the following BibSonomy page. Just try it on your website or blog – it is really easy!

Beate

Feature of the Week: BibSonomy Inside

As started in the last feature of the week we release this week additional information about BibSonomy which will help to integrate BibSonomy into other services like catalog systems of libraries. As discussed the former FOTW BibSonomy relies on different hash keys to find similar publications. The needed information to compute these keys is described here. The page includes a small demo and a java example implementation which is used in BibSonomy to compute the different hash keys. This implementation can be used to compute the same hash keys in other system. Together with the possibility to access the tag cloud of a publication by a hash via BibSonomy's API a nice integration and a community summary of a publication can be integrated into every system.

We think this is a first step for a tighter integration of Bibsonomy and libraries catalog systems or any other similar system.

Feature of the Week: Tags for Resources

If you're using the API and are interested in all tags users have assigned to a specific resource, you now have the possibility to get the tags by using the API-URL http://www.bibsonomy.org/api/tags with the parameter "resource=[hash]" (together with the parameter "resourcetype"). E.g., a request to /api/tags?resourcetype=bookmark&resource=64122a860e632f5ed0148740e25239e6 gives you all tags for the bookmark with the hash 64122a860e632f5ed0148740e25239e6. This allows you to include all tags related to a resource in your application.

Please note, that you have to use the interHash of a resource (as given in the XML output of the API) for this functionality. Further information can be found in the API documentation.

Note: When using hashes of BibTeX posts from the BibSonomy web interface links (e.g., /bibtex/295b16cdb7731d697234d355fc4e68fe3), one must be careful to pick the "correct" hash. In the links all intra-hashes currently have a "2" and all inter-hashes a "1" prepended and are thus 33 characters long (compare /bibtex/295b16cdb7731d697234d355fc4e68fe3/emanuel with /bibtex/1148dc0c64433f631174b5fc419597dd6). This is to distinguish several types of hashes in the web interface (the "old" intra-hashes which started with a "0" still work!). So when using a hash from the web interface, care must be taken that it is really an intra-hash (e.g., starts with a "2") and before querying the API the preceeding "2" must be removed.
We will shortly extend the API to also accept 33 characters long intra-hashes.

BibSonomy backup client available

A golden rule when working with computers is to never forget a regular backup of important data. For the case of your data (BibTeX entries and bookmarks) stored in BibSonomy, we are investing a lot of effort (regular database backups, replication databases, ...) on the server side to prevent data loss in any form, so basically you can sleep sound while we take care of your backup strategy.

However, the most trustworthy backup remains still the one that I have done on my own and the one that is stored on a computer of my choice. To support you in this task, we are happy to present you our brand new backup tool: It is available for download here (check the latest versio of the file bibsonomy-userbackup.zip):

https://gforge.cs.uni-kassel.de/frs/?group_id=78


It is written in Java and is based on our API. It basically

• downloads all your Bibtex and bookmark data (or the public data of other users) to your computer
• saves all data in a zip file
• enables to restore the backed-up data from this zip file back into the BibSonomy account.

To test it out, follow these steps:

1. unzip the .jar-file contained in the downlodable file (see above)
2. if you do not yet have an API-key to access our API, you can obtain one here
3. start the application on the command line by typing java -jar bibsonomy-userbackup.jar (Java needs to be installed on your system at this point)
4. follow the instructions to back-up or restore your BibSonomy data

Please note that in the current status, all data is backed up except private PDF documents and private notes. We are working on including these as well, and we hope that the tool is already useful for you in its current status. If you have questions or comments, don't hesitate to contact us.

Best,
Dominik

Organize your view on collected resources

BibSonomy allows you now to navigate faster through your bookmarks and bibtex entries (see above option in the figure). A new feature is now integrated which offers you to navigate besides going to the next side also to the last side of your bookmark and bibtex collection.

A second option is pointed out at the bottom on the figure. You can choose the numbers of entries. The range is 5-10-20-50-100 and is also valid for bibtex entries.

Best, your BibSonomy Team

Feature of the Week: Using the Webservice with Python

Since more and more users want to use BibSonomy's Webservice, we put together an example how to retrieve posts programmatically - everything with Python. You can have a look at the script here.

If you want to get a list of your publications you would do this:

bibsonomy = BibSonomy('YOUR_USERNAME', 'YOUR_APIKEY')
posts = bibsonomy.getPosts('bibtex')
# do something with the posts...
for post in posts:
print post.resource.title

The last two lines iterate over the retrieved posts and display the title for every post. Have a look at the various classes (Post, BibTex, Tag, etc.) for more information about available attributes.

If you consider using the script and develop it futher, we would be happy to hear from you.

Feature of the (last) Week: Multilanguage Support

As researchers and students of different nationalities work with BibSonomy, its new release starts with integrating multilingual features.
The following pages now include English and German language support:

• homepage (/)
• User-Pages (/user/USERNAME)
• Group-Pages (/group/GROUPNAME)
• Group-Tag-Pages (/group/GROUPNAME/TAGNAME)
  

We would like to extend this offer in the future. If you are proficient in a language other than German or English, you may contribute to this feature by helping us translating the above pages.

Beate