[Corpora-List] 2nd CfP: Mining Track @ Social Book Search Lab 2016, CLEF 2016

Toine Bogers toine at hum.aau.dk
Sun Mar 20 22:22:09 CET 2016


Apologies for cross-posting.

###################################################

CALL FOR PARTICIPATION

Mining Track @ CLEF 2016

(Part of Social Book Search Lab)

Please visit the mining track web page: http://social-book-search.humanities.uva.nl/#/mining

CLEF conference: 5-8 september 2016, Évora, Portugal CLEF website: http://clef2016.clef-initiative.eu/

###################################################

### OVERVIEW ###

The Mining Track @ CLEF 2016 is a shared task that wishes to explore how Natural Language Processing tools and techniques can be used to support book search. It consists of two specific tasks: (1) a classsification task, where the goal is to automatically detect book requests in forum threads; and (2) a book linking task, with the aim of automatically linking books mentioned in forum posts to the proper book metadata records.

The Mining Track is one of the three track of the Social Book Search (SBS) Lab, which investigates book search in scenarios where users search with more than just a query, and look for more than objective metadata. Real-world information needs are generally complex, yet almost all research focuses instead on either relatively simple search based on queries or recommendation based on profiles. The goal is to research and develop techniques to support users in complex book search tasks. 2016 marks the first year that the Mining Track is running as part of SBS.

==Tasks==

Users with search needs that are difficult to express as a search engine query often turn to online forums to get recommendations from others users. The goal of both tasks in the Mining Track is therefore to support social book search by investigating how systems can automatically identify book search requests in online forums and how they can automatically detect and link the books mentioned in these online discussions. The tasks are:

1. Classification -- To identify which threads on online forums are book search requests. That is, given a forum thread with one or more posts, the system should determine whether the opening post contains a request for book suggestions (i.e., binary classification of opening posts as requests or non-requests) 2. Linking -- To recognize book titles in posts (responses) and link them to their unique book ID (and thereby book metadata record). The task is not to mark the entity mention in the post text, but to label the post with the IDs of the mentioned books. That is, the system does not have to identify the exact phrase that refers to book, but only has to identify which book(s) are mentioned.

For more information about the tasks, data format, and evaluation procedure, see http://social-book-search.humanities.uva.nl/#/mining

### SCHEDULE ###

- 25 January: Mining Track training data released - 15 April: Mining test data released - 20 April: Mining run submission deadline

End of CLEF Evaluation Cycle: May 4, 2016 Submission of Participant Papers [CEUR-WS]: May 25, 2016 Notification of Acceptance Participant Papers [CEUR-WS]: June 17, 2016

### HOW TO PARTICIPATE ###

Register at CLEF 2016: http://clef2016-labs-registration.dei.unipd.it/

For registered participants, the training data including book metadata and forum threads are available here: http://social-book-search.humanities.uva.nl/#/data

### ORGANISERS OF MINING TRACK ###

Toine Bogers (Aalborg University Copenhagen) Iris Hendrickx (Radboud University) Marijn Koolen (University of Amsterdam) Suzan Verberne (Radboud University)

### CONTACT ###

For questions regarding the Mining Track, please contact Suzan Verberne (s.verberne at let.ru.nl) or Toine Bogers (toine at hum.aau.dk).



More information about the Corpora mailing list