Information Management and Preservation

MUCKE - News

• Our survey article on Credibility has been published in the 9th Volume of Foundations and Trends in Information Retrieval. Credibility in Information Retrieval
• Deliverable 6.1 is available here
• Deliverable 5.1 is available here
• Deliverable 6.2 is available here
• Deliverable 4.2 is available here
• our CBMI paper is available online: here
• Deliverable 5.3 is available here
• Deliverable 5.2 is available here
• Deliverable 3.3 is available here
• CEA's paper at Multimedia Modelling 2015 received 'best paper award'
• Deliverable 3.2 is available here
• Deliverable 4.1 is available here
• Deliverable 2.3 is available here
• Deliverable 2.2 is available here
• Deliverable 2.1 is available here
• The first protoype of the MUCKE framework was demo-ed at ICMR 2014. The source code is available on GitHub under GPL.
• MUCKE Deliverable 3.1 (Credibility Models for Multimedia Streams) is available here.
• MUCKE and CEA submitted runs to the Retrieving Diverse Social Images Task in MediaEval and came up 3rd and 7th in the final results
• CEA participated in the MediaEval 2013 Placing Task and ranked first of seven participants.
• A new article appeared: Building Specialized Bilingual Lexicons Using Large-Scale Background Knowledge by Dhouha Bouamor, Adrian Popescu, Nassredine Semmar, and Pierre Zweigenbaum in EMNLP 2013 (Seattle, USA)
• Deliverable 1.2 (New Data Collected and Associated Report) is available for download here.
• Deliverable 6.3 (Report on resource sharing framework) is available for download here.
• Alexandru Ginsca (CEA) has worked on the Sentiment Analysis in Twitter track of SemEval and his run was ranked 5th out of 29 participant groups to the Task A: Contextual Polarity Disambiguation. With some adaptations, the method he developed will be useful for credibility estimation in the project.
  More recently, Adrian Popescu (CEA) participated to CLEF CHiC in order to evaluate multilingual implementation of ESA in two settings: ad-hoc retrieval and semantic enrichment. The team ranked 2nd out of 7 participants for ad-hoc retrieval and 1st out of 2 for semantic enrichment.
• Deliverable 1.1 is available.
• The TU Wien team is glad to welcome Ralf Bierig as the new post-doc on the project.
• Internal wiki now online.
• We're looking for a post-doc. Here is the job description.

MUCKE - Abstract

Web3.0 has already appeared in the public vocabulary over 5 years ago. While its definition remains unclear, what has become clear in the last half decade is that the web has become a support for social media. Directly from cameras, phones, tablets or computers, users are pushing multimedia data towards their peers and the world at large. MUCKE addresses this stream of multimedia social data with new and reliable knowledge extraction models designed for multilingual and multimodal data shared on social networks. It departs from current knowledge extraction models, which are mainly quantitative, by giving a high importance to the quality of the processed data, in order to protect the user from an avalanche of equally topically relevant data. It does so using two central innovations: automatic user credibility estimation for multimedia streams and adaptive multimedia concept similarity. Credibility models for multimedia streams are a highly novel topic, which will be cast as a multimedia information fusion task and will constitute the main scientific contribution of the project. Adaptive multimedia concept similarity departs from existing models by creating a semantic representation of the underlying corpora and assigning a probabilistic framework to them. The utility of these two innovations will be demonstrated in an image retrieval system. Extensive evaluation will be performed in order to assess the reliability of the extracted knowledge against representative datasets. Additionally, a new, shared evaluation task focused on user credibility estimation will be proposed. The two core innovations rely on innovative text processing, image processing and fusion methods. Text processing will concentrate on tasks such as word sense disambiguation, concept recognition and anaphora resolution. Image processing will include parsimonious content description, large scale concept detection and detector robustness. Multimedia fusion will focus on a flexible combination of text and image modalities based on a probabilistic framework. All proposed methods will be designed to take advantage of the structural properties of the social networks. Particular focus will be placed on the proposition of scalable algorithms, which cope with large-scale, heterogeneous data.