Summer, 2022
We are proud to present MedDistant19, a new accurate benchmark for broad-coverage biomedical relation extraction at Coling-2022. For more details see https://arxiv.org/abs/2204.04779
Summer, 2021
A number of new research projects have started: XAINES, where we explore explaining AI with narratives. CORA4NLP which is about Co(n)textual Reasoning and Adaptation for Natural Language Processing.
Our linguistic pipeline Munderline has been intgerated into the European Language Grid, cf. link here.
Winter, 2020
Our paper on "LowFER: Low-rank Bilinear Pooling for Link Prediction has been presented at ICML-2020. See also its online presentation.
Summer, 2020
Our paper on "Towards End-to-End Multilingual Question Answering" has been published in the journal Information Systems Frontiers, February, 2020.
Winter, 2019
We successfully participated at CLEF eHealth 2019
in the Multlingual Information Extraction task with our system called
It was also fun participating at FEVER 2.0 (Fact Extraction and Verification) shared task. Our system called Dominiks achieved very good results. See more details here.
Summer, 2019
Our work on "Integrating Heterogeneous Sources for Predicting Question Temporal Anchors across Yahoo! Answers" has been accepted by the journal Information Fusion and is available online and with 50 days' free access.
Summer semster, 2018
I am offering this summer semester a seminar about
April, 2018
New projects at start 2018 about Question&Answering and Deep Learning funded by BMBF and Nuance ! More information is comming soon!
June, 2017
Follow our new project iRead: personalised reading apps for primary school children from ResearchGate and enjoy !
March, 2016
Our newest work "Context-Aware Semantic Classification of Search Queries for Browsing Community Question-Answering Archives" has been published in Knowledge-Based Systems. You can retrieve a copy until March, 25th, from here.
We induce semantic classes of search queries (e.g., “rib pain after ovulation” and “iron oxide household”) by means of contextual information of search sessions. We present Big Data stidies on the basis of an automatically built corpus where we integrate Yahoo! Search query logs with Yahoo! Answers pages.
November, 2015
We have participated in the new TREC 2015 Live Question Answering track (LiveQA). The goal was to automatically answer questions from actual web users of Yahoo Answers in real-time and real-life, i.e., these are fresh new questions basically about anything and the candidate answers cannot easily be retrieved from some ready-made storage or so. For more information see this short post about the competition on the official Yahoo Labs tumblr blog and the official Trec Live QA web page.
September, 2015
New data set about "Customer Interaction Data of German Emails and Online Requests" available from here. It is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
August, 2014
NLG Idol Papers award for our Coling 1992 paper about "Self-Monitoring with Reversible Grammars" presented by Michael White at the 8th International Natural Language Generation Conference (INLG-14), Philadelphia, PA, USA 19.June – 21.June.
July, 2014
A new release EOP 1.1.3 of the Excitement Open Platform for Entailment is available !
June, 2014
Our newest work on Category-specific Models for Ranking Effective Paraphrases in Community Question Answering appears in Journal Expert Systems with Applications and is now online on Science Direct. We present experiments with Big Data from Yahoo! Answers and Yahoo! Search logs and conduct experiments on fine-grained question categories from Yahoo! Answers.
Before June, 2012
We have manually translated parts of the RTE-3 development data from English to German. You can download the data from here.
I am proud to announce that the new EC funded project "EXCITEMENT: EXploring Customer Interactions through Textual EntailMENT" has been launched. Get more information from the official EC project Web page and from here !