Information Extraction and Linking in a Retrieval Context

Marie-Francine Moens and I will give a tutorial at ECIR 2009 on using the results of information extraction and linking for retrieval systems. The tutorial's main goal is to give the participants a clear and detailed overview of content modeling approaches and tools, and the integration of their results into ranking functions. A small set of integrated and interactive exercises will sharpen the understanding by the audience. By attending the tutorial, attendants will:

  • Acquire an understanding of current information extraction, topic modeling and entity linking techniques;
  • Acquire an understanding of ranking models in information retrieval;
  • Be able to integrate the (probabilistic) content models into the ranking models;
  • Be able to choose a model for retrieval that is well-suited for a particular task and to integrate the necessary content models.

The tutorial includes several motivating examples and applications among which are expert search using output from named entity tagging, connecting names to faces in videos for person search using output from named entity tagging and face detection, video search using output from concept detectors, and spoken document retrieval using speech lattices and posterior probabilities of recognized words. The examples will be combined in a larger case study: Retrieval of news broadcast video.

[download pdf]

More info at the ECIR tutorial page.

DIR industry talk by Rene van Erk

Rene van Erk is member of the European management team of Wolters Kluwer where he is responsible for all Product – and Business Development. In this role, his key responsibility is to optimize the WK portfolio for maximum growth, meaning: a) M&A focus: Responsible for identifying acquisition opportunities with a good strategic portfolio b) Leading Innovation: Overall responsible for Product Development & Product Management moving WK from content provider to information solutions provider c) Leading our Online and Software Businesses: Wolters Kluwer currently owns around 30 Software Development Companies and leading Online properties across Europe. Rene will talk about: Communities and Workflow: Driving Information Consumption

More info at: DIR 2009

SIKS PhD Career Day

Imagine. After four (or more) years of hard labour you finally got what you wished for. You have been granted a PhD! Congratulations! … Now what!? During your PhD you may have wondered:

– Which career options do I have after my PhD?
Currently, there are around 200 SIKS PhD alumni. Around 40% of the graduated SIKS PhD's now work in academia. Others work in companies on subjects that are either related or unrelated to their PhD subject, whereas some start their own company. What options do you have?

– What does my dream job look like?
How does a day in the life of a university professor differ from one of a consultant or an entrepreneur? Each job requires different skills and offers different challenges. How can you find out which job suits you best?

– How do I get it?
So you have a long list of publications about the meta-heuristic complexity of recursive retrieval of ubiquitous agent mining in Prolog. Who wouldn't want an expert in this? What else do the various jobs want in a PhD and how can you prepare for that? How does a PhD compare with a MSc with four years working experience? How can you plan your career to get that dream job?

The SIKS PhD Career Day aims at helping you answer these questions. Different people – from academia, companies and research institutes – will give their personal and professional perspective on the three central questions. During this intensive, highly interactive course you will not only listen to presentations but you also have the opportunity to talk to (research) institutes and companies in private. This course is intended for SIKS PhD students in every stage of their PhD, but may be especially interesting for those in a later stage. But also SIKS-alumni are cordially invited to participate.

More info at SIKS.

PF/Tijah facts and figures

The MultimediaN project will finish later this year, and the MultimediaN board asks for the “economic impact” of PF/Tijah.

  • In 2008, the PF/Tijah web site was visited 1,885 times, 6,284 page views in total.
  • During that period, MonetDB/XQuery was downloaded 75 times via the PF/Tijah site. In total MonetDB/XQuery, including PF/Tijah, was downloaded over 2000 times in 2008.

Go to: PF/Tijah.

Call for Participation: Dutch-Belgian IR workshop

The 9th Dutch-Belgian Information Retrieval Workshop (DIR) will take place at the University of Twente, Enschede, the Netherlands, on February 2-3, 2009. The workshop is organised under the auspices of the Working Community on Information Sciences (WGI) and the School for Information and Knowledge Systems (SIKS). The workshop is sponsored by NWO, CTIT and the University of Twente.

Keynote Speaker: prof. Gerhard Weikum (MPII, Saarbruecken, Germany) Gerhard Weikum is Research Director at the Max-Planck Institute for Informatics (MPII) in Saarbruecken, Germany, where he is leading the department on databases and information systems. Prof. Weikum is ACM fellow and a renowned expert in the field of Databases. He received the VLDB 10-Year Achievement Award in 2002. Since then, he focused on several information retrieval problems such as peer-to-peer search, search efficiency, and database and search integration, resulting in for instance 6 full papers at the last SIGIR conferences.

Due to the cooperation between SIKS, SIKS-members can participate without paying entrance fee. Especially PhD-students working on the SIKS-focus “Web based Information Systems” and “Data management, Storage and Retrieval” are strongly encouraged to participate. There is a limited number of registrations for SIKS members. Registrations will be honored in a first come-first serve manner.

Participants of DIR are invited to the Annual Meeting of the Werkgemeenschap Informatieschap which is directly following DIR on Tuesday.

Saving and Accessing the Old IR Literature

SIGIR presents the first results of a project to digitize the older literature in the information retrieval field. So far 14 of the old reports, such as the Cranfield reports and the SMART reports have been scanned, along with Karen Sparck Jones’s Information Retrieval Experiment book. The PDF versions of these are available from the SIGIR Digital Museum of Information Retrieval Research, that provides room for exhibits of historic interest, and allows searching of the material using the PF/Tijah XML search system. The complete library is available for download on request. Requests can be directed to the SIGIR Information Director by sending an email to infodir_sigir@acm.org.

[download pdf]