IM2 Technology transfer event, Idiap research institute, Switzerland2013
RecSys Workshop on Recommendation Utility Evaluation, Dublin, Ireland 2012
Computer Speech and Language, Language Resources and Evaluation (LREV), PLOS ONE, Briefings in BioInformatics, Journal of Biomedical Semantics, BMC Bioinformatics, Data & Knowledge Engineering, Neural Networks, Journal of the American Medical Informatics Association, Drug Safety
Current table similarity measures depend on simple models of table metadata, structure, and content. They are designed mainly for relational tables, and cannot be easily applied to tables with other structures, such as matrix tables where both rows and columns are represented by attributes and values. Moreover, they rely on frequency-based methods which are not sufficient to capture the semantics of table elements. The main objective of this projetc is to research methods that bring more semantics to table similarity measures.
PatSeg (Biomedical Patent Segmentation), Bayer AG 2017-2018
Patents are a key source of information for most industries. They are often the first channel of publication of new ideas, innovations, and technologies. Due to the steep growth of the number of published patent applications and due to the typical length of patent applications, it has become extremely challenging to keep track of novel innovations. The aim of this project is the use of information extraction methods to automatically extract relevant knowledge from patents.
Biomedical patent mining is particularly essential due to the high economic importance of pharmaceutical findings. Biomedical patent mining refers to the development of methods for recognition of biomedical named entities, normalizing them into a database identifier, and identifying relation among different biomedical named entities. Here, the goal is designing novel text mining methods using deep learning for patents which are typically much lengthy, difficult to understand and have a lower word density.
Multilingual conversations as an effective way of knowledge transfer among parties with diverse languages are highly influenced by the progress in automatic machine translation. The main objective of this project is to research methods that provide more contextual information for translation system using conversational content.
REMUS, Hasler foundation2012-2014
People are surrounded by an unprecedented wealth of information.
Access to it depends on the availability of suitable search engines.
The goal of this project is to study methods that refine users' queries using their current activities.
IM2 NCCR, SNSF2011-2014
Human beings face an unexpectedly high volume of information, available as documents,
databases, or multimedia resources. However, humans often do not initiate a search to access new information, because their current activity does not allow them to do so, or because they are not aware that relevant information is available. In this project, a set of novel approaches are studied to model users' information needs based on their current activities and retrieve potentially useful documents.
M. Habibi, A. Rheinlaender, W. Thielemann, R. Adams, P. Fischer, S. Krolkiewicz, D. L. Wiegandt & U. Leser (2020) - "PatSeg: a Sequential Patent Segmentation Approach", Big Data Research.
L. Weber, J. Münchmeyer, T. Rocktäschel, M. Habibi, & U Leser (2020) - HUNER: Improving Biomedical NER with Pretraining, Bioinformatics.
M. Habibi, L. Weber, M. Neves, D. L. Wiegandt & U. Leser (2017) - Deep Learning with Word Embeddings Improves Biomedical Named Entity Recognition, Bioinformatics.
M. Habibi, D. L. Wiegandt, F. Schmedding & U. Leser (2016) - Recognizing Chemicals in Patents – a Comparative Analysis, Journal of Cheminformatics. [pdf][bibtex]
M. Habibi, P. Mahdabi & A. Popescu-Belis (2016) - Question Answering in Conversations: Query Refinement Using Contextual and Semantic Information, Data & Knowledge Engineering.[pdf][bibtex]
M. Habibi & A. Popescu-Belis (2015) - Keyword Extraction and Clustering for Document Recommendation in Conversations, IEEE/ACM Transactions on Audio, Speech and Language Processing (TASLP).[pdf][bibtex][code]
M. Habibi, H. Sameti & H. Setareh (2010) - On-Line Learning of a Persian Spoken Dialogue System Using Real Training Data. Journal of Advances In Computer Research.[pdf][bibtex]
M. Habibi, D. L. Wiegandt, F. Schmedding & U. Leser (2016) - Performance of Gene Name Recognition Tools on Patents. Semantic Mining in Biomedicine (SMBM).[pdf][bibtex]
M. Habibi, & A. Popescu-Belis (2015) - Query Refinement Using Local Context from Conversations: a Method and a Resource for its Evaluation. International Conference on Application of Natural Language to Information Systems (NLDB) (Nominated as one of the best papers).[pdf][bibtex]
M. Habibi & A. Popescu-Belis (2014) - Enforcing Topic Diversity in a Document Recommender for Conversations. International Conference on Computational Linguistics (Coling).[pdf][bibtex]
C. Bhatt, N. Pappas, M. Habibi & A. Popescu-Belis (2014) - Multimodal Reranking of Content-based Recommendations for Hyperlinking Video Snippets. ACM International Conference on Multimedia Retrieval (ACM ICMR), special session on User-centric Video Search and Hyperlinking.[pdf][bibtex]
M. Habibi & A. Popescu-Belis (2013) - Diverse Keyword Extraction from Conversations. Annual Meeting of the Association for Computational Linguistics (ACL).[pdf][bibtex]
C. Bhatt, N. Pappas, M. Habibi & A. Popescu-Belis (2013) - Idiap at MediaEval 2013: Search and Hyperlinking Task. MediaEval 2013 Workshop.[pdf][bibtex]
C. Bhatt, A. Popescu-Belis, M. Habibi, S. Ingram, F. McInnes, S. Masneri, N. Pappas & O. Schreer (2013) - Multi-factor Segmentation for Topic Visualization and Recommendation: the MUST-VIS System. ACM International Conference on Multimedia (MM 2013), Grand Challenge Solutions.[pdf][bibtex]
M. Habibi & A. Popescu-Belis (2012) - Using Crowdsourcing to Compare Document Recommendation Strategies for Conversations. ACM RecSys Workshop on Recommendation Utility Evaluation: Beyond RMSE (RUE).
, S. Rahbar & H. Sameti (2010) - Divided POMDP Method for Complex Menu Problems in Spoken Dialogue Systems. Spoken Language Technology Workshop (SLT).
M. Habibi, H. Sameti & H. Setareh (2010) - On-Line Learning of a Persian Spoken Dialogue System Using Real Training Data. International Conference on Information Sciences, Signal Processing and their Applications (ISSPA).
M. Habibi, N. Pappas & A. Popescu-Belis (2017) - Topic and Sentiment in Phrase-Based Statistical Machine Translation. Idiap Research Institute.
A. Popescu-Belis, M. Habibi, P. N. Garner & N. Li (2017) - From Research to Reality: Evaluation of a Single-Computer Real-Time LVCSR System for Speech-Based Retrieval. Idiap Research Institute.
PhD thesis: Modeling Users’ Information Needs in a
Document Recommender for Meetings. EPFL (2015).
Master thesis: Reinforcement Learning in Spoken Dialogue Management
Systems. Sharif University of Technology (2010).
Collaboration in “Beyond the Exome” funding proposal, DFG Foundation, Germany 2019
Co-author in “TabSim” funding proposal, DFG Foundation, Germany 2017