Paper Published in Neurocomputing

We are happy to announce that our paper “Trans4E: Link Prediction on Scholarly Knowledge Graph” has been published in Neurocomputing. Neurocomputing publishes articles describing recent fundamental contributions in the field of neurocomputing. Neurocomputing theory, practice and applications are the essential topics being covered.

Here is the abstract and the link to the paper:

Trans4E: Link Prediction on Scholarly Knowledge Graph
By Mojtaba Nayyeri, Gokce Muge Cil, Sahar Vahdati, Francesco Osborne, Mahfuzur Rahman, Simone Angioni,, Angelo Salatino, Diego Reforgiato Recupero, Nadezhda Vassilyeva, Enrico Motta and Jens Lehmann.
Abstract The incompleteness of Knowledge Graphs (KGs) is a crucial issue affecting the quality of AI-based services. In the scholarly domain, KGs describing research publications typically lack important infor-mation, hindering our ability to analyse and predict research dynamics. In recent years, link prediction approaches based on Knowledge Graph Embedding models became the first aid for this issue. In this work, we present Trans4E, a novel embedding model that is particularly fit for KGs which include N to M relations with N≫M. This is typical for KGs that categorize a large number of entities (e.g., re-search articles, patents, persons) according to a relatively small set of categories. Trans4E was applied on two large-scale knowledge graphs, the Academia/Industry DynAmics (AIDA) and Microsoft Academic Graph (MAG), for completing the information about Fields of Study (e.g., ’neural networks’,’machine learning’, ’artificial intelligence’), and affiliation types (e.g., ’education’, ’company’, ’gov-ernment’), improving the scope and accuracy of the resulting data. We evaluated our approach against alternative solutions on AIDA, MAG, and four other benchmarks (FB15k, FB15k-237, WN18, and WN18RR). Trans4E outperforms the other models when using low embedding dimensions and obtains competitive results in high dimensions.

Papers Accepted At ESWC 2021

We are very pleased to announce that our group got four papers accepted for presentation at ESWC2021. The ESWC is a major venue for discussing the latest scientific results and technology innovations around semantic technologies. Building on its past success, ESWC is seeking to broaden its focus to span other relevant related research areas in which Web semantics plays an important role. The goal of the Semantic Web is to create a Web of knowledge and services in which the semantics of content is made explicit and content is linked to both other content and services allowing novel applications to combine content from heterogeneous sites in unforeseen ways and support enhanced matching between users needs and content. This network of knowledge-based functionality will weave together a large network of human knowledge, and make this knowledge machine-processable to support intelligent behaviour by machines. Creating such an interlinked Web of knowledge which spans unstructured text, structured data (e.g. RDF) as well as multimedia content and services requires the collaboration of many disciplines, including but not limited to: Artificial Intelligence, Natural Language Processing, Databases and Information Systems, Information Retrieval, Machine Learning, Multimedia, Distributed Systems, Social Networks, Web Engineering, and Web Science.

Here are the abstracts and the links to the papers:

  • Grounding Dialogue Systems via Knowledge Graph Aware Decoding with Pre-trained Transformers
    By Debanjan Chaudhuri, Md Rashad Al Hasan Rony, and Jens Lehmann.
    Abstract Generating knowledge grounded responses in both goal and non-goal oriented dialogue systems is an important research challenge. Knowledge Graphs (KG) can be viewed as an abstraction of the real world, which can potentially facilitate a dialogue system to produce knowledge grounded responses. However, integrating KGs into the dialogue generation process in an end-to-end manner is a non-trivial task. This paper proposes a novel architecture for integrating KGs into the response generation process by training a BERT model that learns to answer using the elements of the KG (entities and relations) in a multi-task, end-to-end setting. The k-hop subgraph of the KG is incorporated into the model during training and inference using Graph Laplacian. Empirical evaluation suggests that the model achieves better knowledge groundedness (measured via Entity F1 score) compared to other state-of-the-art models for both goal and non-goal oriented dialogues.
  • Context Transformer with Stacked Pointer Networks for Conversational Question Answering over Knowledge Graphs
    By Joan Plepi,, Endri Kacupaj, Kuldeep Singh, Harsh Thakkar, and Jens Lehmann.
    Abstract Neural semantic parsing approaches have been widely used for Question Answering (QA) systems over knowledge graphs. Such methods provide the flexibility to handle QA datasets with complex queries and a large number of entities. In this work, we propose a novel framework named CARTON, which performs multi-task semantic parsing for handling the problem of conversational question answering over a large-scale knowledge graph. Our framework consists of a stack of pointer networks as an extension of a context transformer model for parsing the input question and the dialog history. The framework generates a sequence of actions that can be executed on the knowledge graph. We evaluate CARTON on a standard dataset for complex sequential question answering on which CARTON outperforms all baselines. Specifically, we observe performance improvements in F1-score on eight out of ten question types compared to the previous state of the art. For logical reasoning questions, an improvement of 11 absolute points is reached.
  • ParaQA: A Question Answering Dataset with Paraphrase Responses for Single-Turn Conversation
    By Endri Kacupaj, Barshana Banerjee, Kuldeep Singh, and Jens Lehmann.
    Abstract This paper presents ParaQA, a question answering (QA) dataset with multiple paraphrased responses for single-turn conversation over knowledge graphs (KG). The dataset was created using a semi-automated framework for generating diverse paraphrasing of the answers using techniques such as back-translation. The existing datasets for conversational question answering over KGs (single-turn/multi-turn) focus on question paraphrasing and provide only up to one answer verbalization. However, ParaQA contains 5000 question-answer pairs with a minimum of two and a maximum of eight unique paraphrased responses for each question. We complement the dataset with baseline models and illustrate the advantage of having multiple paraphrased answers through commonly used metrics such as BLEU and METEOR. The ParaQA dataset is publicly available on a persistent URI for broader usage and adaptation in the research community.
  • A Virtual Knowledge Graph for Enabling DefectTraceability and Customer Service Analytics
    By Nico Wilhelm, Diego Collarana, and Jens Lehmann.
    Abstract In this paper, we showcase the implementation of a semantic information model and a virtual knowledge graph at ZF Friedrichshafen AG company, with two main goals in mind: 1) integration of heterogeneous data sources following a pay-as-you-go approach; and the 2) combination core domain concepts from ZF’s production line with meta-data of its internal data sources. We employ the developed semantic information model in two use cases, defect traceability and customer service, demonstrating and discussing the benefits and opportuni-ties provided by following an agile semantic virtual integration approach.

Paper Accepted At ECIR 2021

We are very pleased to announce that our group got a paper accepted for presentation at ECIR 2021. The ECIR conference is the premier European forum for the presentation of new research results in the broadly conceived area of Information Retrieval (IR), and has a strong focus on the active participation of early-career researchers. The General Chairs of ECIR 2021 invite researchers, academics, students, and industry leaders working in the field to join us online for a rich program featuring full-paper and poster presentations, system demonstrations, tutorials, workshops, an industry-oriented event, and great social events.

Here is the abstract and the link to the paper:

Pattern-Aware and Noise-Resilient Embedding Models
By Mojtaba Nayyeri, Sahar Vahdati, Emanuel Sallinger, Mirza Mohtashim Alam, Hamed Shariat Yazdi and Jens Lehmann.
Abstract Knowledge Graph Embeddings (KGE) have become an important area of Information Retrieval (IR), in particular as they provide one of the state-of-the-art methods for Link Prediction. Recent work in the area of KGEs has shown the importance of relational patterns, i.e., logical formulas, to improve the learning process of KGE models significantly. In separate work, the role of noise in many knowledge discovery and IR settings has been studied, including the KGE setting. So far, very few papers have investigated the KGE setting considering both relational patterns and noise. Not considering both together can lead to problems in the performance of KGE models. We investigate the effect of noise in the presence of patterns. We show that by introducing a new loss function that is both pattern-aware and noise-resilient, significant performance issues can be solved. The proposed loss function is model-independent which could be applied in combination with different models. We provide an experimental evaluation both on synthetic and real-world cases.

Paper Accepted at NAACL 2021

We are very pleased to announce that our group got a paper accepted for presentation at NAACL21.The North American Chapter of the Association for Computational Linguistics (NAACL) provides a regional focus for members of the Association for Computational Linguistics (ACL) in North America as well as in Central and South America, organizes annual conferences, promotes cooperation and information exchange among related scientific and professional societies, encourages and facilitates ACL membership by people and institutions in the Americas, and provides a source of information on regional activities for the ACL Executive Committee.

Here is the abstract and the link to the paper:

Temporal Knowledge Graph Completion using a Linear Temporal Regularizer and Multivector Embeddings
By Chengjin Xu, Yung-Yu Chen, Mojtaba Nayyeri, and Jens Lehmann.
Abstract Representation learning approaches for knowledge graphs have been mostly designed for static data. However, many knowledge graphs involve evolving data, e.g., the fact (The President of the United States is Barack Obama) is valid only from 2009 to 2017. This introduces important challenges for knowledge representation learning since the knowledge graphs change over time. In this paper, we present a novel time-aware knowledge graph embedding approach, TeLM, which performs 4th-order tensor factorization of a Temporal knowledge graph using a Linear temporal regularizer and Multivector embeddings. Moreover, we investigate the effect of the temporal dataset’s time granularity on temporal knowledge graph completion. Experimental results demonstrate that our proposed models trained with the linear temporal regularizer achieve state-of-the-art performances on link prediction over four well-established temporal knowledge graph completion benchmarks.

Paper Accepted at IDA 2021

We are very pleased to announce that our group got a paper accepted for presentation at IDA2021. Advancing Intelligent Data Analysis requires novel, potentially game-changing ideas. IDA’s mission is to promote ideas over performance: a solid motivation can be as convincing as exhaustive empirical evaluation.

Here is the abstract and the link to the paper:

HORUS-NER: A Multimodal Named Entity Recognition Framework for Noisy Data
By Diego Esteves, José Marcelino,, Piyush Chawla, Asja Fischer,, and Jens Lehmann.
Abstract Recent work based on Deep Learning presents state-of-the-art (SOTA) performance in the named entity recognition (NER) task. However, such models still have the performance drastically reduced in noisy data (eg, social media, search engines), when compared to the formal domain (eg, newswire). Thus, designing and exploring new methods and architectures is highly necessary to overcome current challenges. In this paper, we shift the focus of existing solutions to an entirely different perspective. We investigate the potential of embedding word-level features extracted from images and news. We performed a very comprehensive study in order to validate the hypothesis that images and news (obtained from an external source) may boost the task on noisy data, revealing very interesting findings. When our proposed architecture is used:(1) We beat SOTA in precision with simple CRFs models (2) The overall performance of decision trees-based models can be drastically improved.(3) Our approach overcomes off-the-shelf models for this task.(4) Images and text consistently increased recall over different datasets for SOTA, but at cost of precision. All experiment configurations, data and models are publicly available to the research community at

Paper Accepted at IJCNN 2021

We are very pleased to announce that our group got a paper accepted for presentation at IJCNN 2021. The annual International Joint Conference on Neural Networks (IJCNN) is the flagship conference of the IEEE Computational Intelligence Society and the International Neural Network Society. It covers a wide range of topics in the field of neural networks, from biological neural network modeling to artificial neural computation.

Here is the abstract and the link to the paper:

Multiple Run Ensemble Learning with Low-Dimensional Knowledge Graph Embeddings
By Chengjin Xu, Mojtaba Nayyeri, Sahar Vahdati, and Jens Lehmann.
Abstract Knowledge graphs (KGs) represent world facts in a structured form. Although knowledge graphs are quantitatively huge and consist of millions of triples, the coverage is still only a small fraction of world’s knowledge. Among the top approaches of recent years, link prediction using knowledge graph embedding (KGE) models has gained significant attention for knowledge graph completion. Various embedding models have been proposed so far, among which, some recent KGE models obtain state-of-the-art performance on link prediction tasks by using embeddings with a high dimension (e.g. 1000) which accelerate the costs of training and evaluation considering the large scale of KGs. In this paper, we propose a simple but effective performance boosting strategy for KGE models by using multiple low dimensions in different repetition rounds of the same model. For example, instead of training a model one time with a large embedding size of 1200, we repeat the training of the model 6 times in parallel with an embedding size of 200 and then combine the 6 separate models for testing while the overall numbers of adjustable parameters are same (6*200=1200) and the total memory footprint remains the same. We show that our approach enables different models to better cope with their expressiveness issues on modeling various graph patterns such as symmetric, 1-n, n-1 and n-n. In order to justify our findings, we conduct experiments on various KGE models. Experimental results on standard benchmark datasets, namely FB15K, FB15K-237 and WN18RR, show that multiple low-dimensional models of the same kind outperform the corresponding single high-dimensional models on link prediction in a certain range and have advantages in training efficiency by using parallel training while the overall numbers of adjustable parameters are same.

Paper Accepted at ICSC 2021

We are very pleased to announce that our group got a paper accepted for presentation at IEEE-ICSC 2021. The 15th IEEE International Conference on Semantic Computing (ICSC2021) addresses the derivation, description, integration, and use of semantics (“meaning”, “context”, “intention”) for all types of resource including data, document, tool, device, process and people. The scope of ICSC2021 includes, but is not limited to, analytics, semantics description languages and integration (of data and services), interfaces, and applications.

Here is the abstract and the link to the paper (we also provide a preprint):

Scalable Distributed in-Memory Semantic Similarity Estimation for RDF Knowledge Graphs with DistSim
By Carsten Draschner, Jens Lehmann, and Hajira Jabeen.
Abstract In this paper, we present DistSim, a Scalable Distributed in-Memory Semantic Similarity Estimation framework for Knowledge Graphs. DistSim provides a multitude of state-of-the-art similarity estimators. We have developed the Similarity Estimation Pipeline by combining generic software modules. For large scale RDF data, DistSim proposes MinHash with locality sensitivity hashing to achieve better scalability over all-pair similarity estimations. The modules of DistSim can be set up using a multitude of (hyper)-parameters allowing to adjust the tradeoff between information taken into account, and processing time. Furthermore, the output of the Similarity Estimation Pipeline is native RDF. DistSim is integrated into the SANSA stack, documented in scala-docs, and covered by unit tests. Additionally, the variables and provided methods follow the Apache Spark MLlib name-space conventions. The performance of DistSim was tested over a distributed cluster, for the dimensions of data set size and processing power versus processing time, which shows the scalability of DistSim w.r.t. increasing data set sizes and processing power. DistSim is already in use for solving several RDF data analytics related use cases. Additionally, DistSim is available and integrated into the open-source GitHub project SANSA.

Paper on Knowledge Graph Integration into Transformer Architectures Accepted at ACL21

We are happy to announce that we got a paper accepted for presentation at ACL 2021 (Association for Computational Linguistics). ACL is a premier Natural Language Processing conference. In the paper, we investigate the efficient integration of knowledge graphs into Transformer-based decoder architectures. The approach allows to integrate knowledge graphs into large-scale language models like GPT-2 or GPT-3, which leads to more comprehensive and interesting dialogues with such models.

Here is the pre-print of the accepted paper with its abstract:

Space Efficient Context Encoding for Non-Task-Oriented Dialogue Generation with Graph Attention Transformer
By Fabian Galetzka, Jewgeni Rose, David Schlangen,Jens Lehmann.
Abstract To improve the coherence and knowledge retrieval capabilities of non task-oriented dialogue systems, recent Transformer-based models aim to integrate fixed background context. This often comes in the form of knowledge graphs, and the integration is done by creating pseudo utterances through paraphrasing knowledge triples, added into the accumulated dialogue context. However, the context length is fixed in these architectures, which restricts how much background or dialogue context can be kept. In this work, we propose a more concise encoding for background context structured in form of knowledge graphs, by expressing the graph connections through restrictions on the attention weights. The results of our human evaluation show, that this encoding reduces space requirements without negative effects on the precision of reproduction of knowledge and perceived consistency. Further, models trained with our proposed context encoding generate dialogues that are judged to be more comprehensive and interesting.

Paper Accepted at EACL21

We are happy to announce that we got a paper accepted for presentation at EACL21 (European Chapter of the ACL). The European Chapter of the ACL (EACL) is the primary professional association for computational linguistics in Europe.

Here is the pre-print of the accepted paper with its abstract:

Conversational Question Answering over Knowledge Graphs with Transformer and Graph Attention Networks
By Endri Kacupaj, Joan Plepi, Kuldeep Singh, Harsh Thakkar,Jens Lehmann, and Maria Maleshkova.
Abstract This paper addresses the task of (complex) conversational question answering over a knowledge graph. For this task, we propose LASAGNE (muLti-task semAntic parSing with trAnsformer and Graph atteNtion nEtworks). It is the first approach, which employs a transformer architecture extended with Graph Attention Networks for multi-task neural semantic parsing. LASAGNE uses a transformer model for generating the base logical forms, while the Graph Attention model is used to exploit correlations between (entity) types and predicates to produce node representations. LASAGNE also includes a novel entity recognition module which detects, links, and ranks all relevant entities in the question context. We evaluate LASAGNE on a standard dataset for complex sequential question answering, on which it outperforms existing baseline averages on all question types. Specifically, we show that LASAGNE improves the F1-score on eight out of ten question types; in some cases, the increase in F1-score is more than 20% compared to the state of the art.

Paper Accepted at PAKDD 2021

We are happy to announce that we got a paper accepted for presentation at PAKDD 2021 (Pacific-Asia Conference on Knowledge Discovery and Data Mining). PAKDD is one of the longest established and leading international conferences in the areas of data mining and knowledge discovery. It provides an international forum for researchers and industry practitioners to share their new ideas, original research results, and practical development experiences from all KDD related areas, including data mining, data warehousing, machine learning, artificial intelligence, databases, statistics, knowledge engineering, visualization, decision-making systems, and the emerging applications.

Here is the pre-print of the accepted paper with its abstract:

Loss-aware Pattern Inference: A Correction on the Wrongly Claimed Limitations of Embedding Models
By Mojtaba Nayyeri, Chengjin Xu, Yadollah Yaghoobzadeh, Sahar Vahdati,Mirza Mohtashim Alam,Hamed Shariat Yazdi and Jens Lehmann.
Abstract Embedding knowledge graphs (KGs) into a low dimensional space has become an active research domain which is broadly utilized in many of the AI-based tasks, especially link prediction. One of the crucial aspects is the extent to which a KG embedding model (KGE) is capable to model and infer various relation patterns, such as symmetry/antisymmetry, inversion, and composition. Each embedding model is highly affected in the optimization of embedding vectors by their loss function which consequently affects the inference of relational patterns. However, most existing methods failed to consider this aspect in their inference capability. In this paper, we show that disregarding loss functions results in inaccurate or even wrong interpretation from the capability of the models. We provide deep theoretical investigations of the already existing KGE models on the example of the TransE model. To the best of our knowledge, so far, this has not been comprehensively investigated. We show that by a proper selection of the loss function for training a KGE e.g., TransE, the main inference limitations are mitigated. The provided theories together with the experimental results confirm the importance of loss functions for training KGE models and their performance.