Skip to main content

George Giannakopoulos

National Centre of Scientific Research "DEMOKRITOS", Institute of Informatics and Telecommunications, Post-Doc

Followers

213

Following

5

Co-authors

2

Public Views

I work as a post-doctoral Research Fellow for the Institute of Informatics and Telecommunications of NCSR "Demokritos", Greece (www.iit.demokritos.gr). I am also the co-founder and Chief Technical Officer of SciFY PNPC (www.scify.org), which brings Artificial Intelligence applications and systems to the public as open systems (open source, open hardware), for free.I received my PhD on Automatic Summarization from the Department of Information and Communication Systems Engineering of the University of the Aegean, while working in NCSR "Demokritos". My expertise lies in Natural Language Processing, Machine Learning, Data Mining and the Semantic Web. I am the organizer of the Multilingual Multi-document Summarization Community (MultiLing) and the corresponding workshops (MultiLing 2011, MultiLing 2013, MultiLing 2015). I have participated in various EU projects (OKKAM, SYNC3, NOMAD, SemaGrow, YourDataStories) and have more than 15 years of experience in the IT industry (medical informatics, mobile solutions, IT consulting, Open source projects). I was honored to receive a Best Paper award in 2009 on adaptive systems. I am a member of the European Chapter of the Association of Computational Linguistics and of the Hellenic Artificial Intelligence Society (EETN). In April 2012, I co-founded SciFY (scify.org) to cover the gap between scientific results and everyday life, through open source projects and knowledge sharing.
Supervisors: George Vouros, Panagiotis Stamatopoulos, and Vangelis Karkaletsis

less

InterestsView All (7)

Uploads

Papers by George Giannakopoulos

JedAI 3 : beyond batch, blocking-based Entity Resolution

Extending Database Technology, 2020

JedAI is an open-source toolkit that allows for building and benchmarking thousands of schema-agn... more JedAI is an open-source toolkit that allows for building and benchmarking thousands of schema-agnostic Entity Resolution (ER) pipelines through a non-learning, blocking-based end-to-end workflow. In this paper, we present its latest release, JedAI 3 , which conveys two new end-to-end workflows: one for budgetagnostic ER that is based on similarity joins, and one for budgetaware (i.e., progressive) ER. This version also adds support for pre-trained word or character embeddings and connects JedAI to the Python data analysis ecosystem. Overall, these enhancements provide JedAI with features offered by no other ER tool, especially in the schema-and domain-agnostic context.

Generating Realistic Nanorough Surfaces Using an N-Gram-Graph Augmented Deep Convolutional Generative Adversarial Network

Data mining for predicting gas diffusivity in zeolitic-imidazolate frameworks (ZIFs)

Journal of materials chemistry. A, Materials for energy and sustainability, 2022

Please note that technical editing may introduce minor changes to the text and/or graphics, which... more Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal's standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.

Summary Evaluation: Together We Stand NPowER-ed

Lecture Notes in Computer Science, 2013

ABSTRACT Summary evaluation has been a distinct domain of research for several years. Human summa... more ABSTRACT Summary evaluation has been a distinct domain of research for several years. Human summary evaluation appears to be a high-level cognitive process and, thus, difficult to reproduce. Even though several automatic evaluation methods correlate well to human evaluations over systems, we fail to get equivalent results when judging individual summaries. In this work, we propose the NPowER evaluation method based on machine learning and a set of methods from the family of &quot;n-gram graph&quot;-based summary evaluation methods. First, we show that the combined, optimized use of the evaluation methods outperforms the individual ones. Second, we compare the proposed method to a combination of ROUGE metrics. Third, we study and discuss what can make future evaluation measures better, based on the results of feature selection. We show that we can easily provide per summary evaluations that are far superior to existing performance of evaluation systems and face different measures under a unified view.

Reproducible experiments on Three-Dimensional Entity Resolution with JedAI

Information Systems, Dec 1, 2021

In Papadakis et al. [1], we presented the latest release of JedAI, an open-source Entity Resoluti... more In Papadakis et al. [1], we presented the latest release of JedAI, an open-source Entity Resolution (ER) system that allows for building a large variety of end-to-end ER pipelines. Through a thorough experimental evaluation, we compared a schema-agnostic ER pipeline based on blocks with another schema-based ER pipeline based on similarity joins. We applied them to 10 established, real-world datasets and assessed them with respect to effectiveness and time efficiency. Special care was taken to juxtapose their scalability, too, using seven established, synthetic datasets. Moreover, we experimentally compared the effectiveness of the batch schema-agnostic ER pipeline with its progressive counterpart. In this companion paper, we describe how to reproduce the entire experimental study that pertains to JedAI's serial execution through its intuitive user interface. We also explain how to examine the robustness of the parameter configurations we have selected.

GeoSensor is a novel, open-source system that enriches change detection over satellite images wit... more GeoSensor is a novel, open-source system that enriches change detection over satellite images with event detection over news items and social media content. GeoSensor combines these two orthogonal operations through state-of-the-art Semantic Web technologies. At its core lies the open-source, semantics-enabled Big Data infrastructure developed by the EU H2020 BigDataEurope project. This allows GeoSensor to offer an on-line functionality, despite facing three major challenges of Big Data: Volume (a single satellite image typically occupies a few GBs), Variety (its data sources include two different types of satellite images and various types of user-generated content) and Veracity, as the accuracy of the end result is crucial for the usefulness of our system. We present GeoSensor&amp;amp;amp;#39;s architecture in detail, highlighting the advantages of using semantics for taking the most of the knowledge extracted from news items and Earth Observation products. We also verify GeoSensor&amp;amp;amp;#39;s efficiency through a preliminary experimental study.

A Graph-based Text Similarity Measure That Employs Named Entity Information

Text comparison is an interesting though hard task, with many applications in Natural Language Pr... more Text comparison is an interesting though hard task, with many applications in Natural Language Processing. This work introduces a new text-similarity measure, which employs named-entities' information extracted from the texts and the ngram graphs' model for representing documents. Using OpenCalais as a namedentity recognition service and the JIN-SECT toolkit for constructing and managing n-gram graphs, the text similarity measure is embedded in a text clustering algorithm (k-Means). The evaluation of the produced clusters with various clustering validity metrics shows that the extraction of named entities at a first step can be profitable for the time-performance of similarity measures that are based on the n-gram graph representation without affecting the overall performance of the NLP task.

Detecting and exploiting stability in evolving heterogeneous information spaces

Individuals contribute content on the Web at an unprecedented rate, accumulating immense quantiti... more Individuals contribute content on the Web at an unprecedented rate, accumulating immense quantities of (semi-) structured data. Wisdom of the Crowds theory advocates that such information (or parts of it) is constantly overwritten, updated, or even deleted by other users, with the goal of rendering it more accurate, or up-to-date. This is particularly true for the collaboratively edited, semi-structured data of entity repositories, whose entity profiles are consistently kept fresh. Therefore, their core information that remain stable with the passage of time, despite being reviewed by numerous users, are particularly useful for the description of an entity. Based on the above hypothesis, we introduce a classification scheme that predicts, on the basis of statistical and content patterns, whether an attribute (i.e., name-value pair) is going to be modified in the future. We apply our scheme on a large, real-world, versioned dataset and verify its effectiveness. Our thorough experimental study also suggests that reducing entity profiles to their stable parts conveys significant benefits to two common tasks in computer science: information retrieval and information integration.

Representation models for text classification

Text classification constitutes a popular task in Web research with various applications that ran... more Text classification constitutes a popular task in Web research with various applications that range from spam filtering to sentiment analysis. To address it, patterns of cooccurring words or characters are typically extracted from the textual content of Web documents. However, not all documents are of the same quality; for example, the curated content of news articles usually entails lower levels of noise than the user-generated content of the blog posts and the other Social Media. In this paper, we provide some insight and a preliminary study on a tripartite categorization of Web documents, based on inherent document characteristics. We claim and support that each category calls for different classification settings with respect to the representation model. We verify this claim experimentally, by showing that topic classification on these different document types offers very different results per type. In addition, we consider a novel approach that improves the performance of topic classification across all types of Web documents: namely the n-gram graphs. This model goes beyond the established bag-of-words one, representing each document as a graph. Individual graphs can be combined into a class graph and graph similarities are then employed to position and classify documents into the vector space. Accuracy is increased due to the contextual information that is encapsulated in the edges of the n-gram graphs; efficiency, on the other hand, is boosted by reducing the feature space to a limited set of dimensions that depend on the number of classes, rather than the size of the vocabulary. Our experimental study over three largescale, real-world data sets validates the higher performance of n-gram graphs in all three domains of Web documents.

Three-dimensional Entity Resolution with JedAI

Information Systems, Nov 1, 2020

Abstract Entity Resolution (ER) is the task of detecting different entity profiles that describe ... more Abstract Entity Resolution (ER) is the task of detecting different entity profiles that describe the same real-world objects. To facilitate its execution, we have developed JedAI , an open-source system that puts together a series of state-of-the-art ER techniques that have been proposed and examined independently, targeting parts of the ER end-to-end pipeline. This is a unique approach, as no other ER tool brings together so many established techniques. Instead, most ER tools merely convey a few techniques, those primarily developed by their creators. In addition to democratizing ER techniques, JedAI goes beyond the other ER tools by offering a series of unique characteristics: (i) It allows for building and benchmarking millions of ER pipelines. (ii) It is the only ER system that applies seamlessly to any combination of structured and/or semi-structured data. (iii) It constitutes the only ER system that runs seamlessly both on stand-alone computers and clusters of computers — through the parallel implementation of all algorithms in Apache Spark. (iv) It supports two different end-to-end workflows for carrying out batch ER (i.e., budget-agnostic), a schema-agnostic one based on blocks, and a schema-based one relying on similarity joins. (v) It adapts both end-to-end workflows to budget-aware (i.e., progressive) ER. We present in detail all features of JedAI , stressing the core characteristics that enhance its usability, and boost its versatility and effectiveness. We also compare it to the state-of-the-art in the field, qualitatively and quantitatively, demonstrating its state-of-the-art performance over a variety of large-scale datasets from different domains. The central repository of the JedAI ’s code base is here: https://github.com/scify/JedAIToolkit . A video demonstrating the JedAI’s Web application is available here: https://www.youtube.com/watch?v=OJY1DUrUAe8 .

Hybridizing AI and Domain Knowledge in Nanotechnology: the Example of Surface Roughness Effects on Wetting Behavior

In this paper, we propose a scheme for the hybridization of domain modeling and theoretical knowl... more In this paper, we propose a scheme for the hybridization of domain modeling and theoretical knowledge in nanotechnology with Artificial Intelligence(AI) techniques and evaluate the success of its application to predict the relationship between nanosurface morphology and wettability. We utilize domain knowledge consisting of two parts. The first part is a mathematical modeling based on the inverse Fourier transform for the generation of rough surfaces with Gaussian or non-Gaussian height distributions, characterized by their first moments (Rms, skewness, kurtosis) and the correlation lengths along x and y-axes. The second part lies in the assumption that the Wenzel scenario for wetting of rough surfaces holds where the critical parameter for contact angle determination is the roughness ratio , defined as the ratio of true (active) area of the solid surface to the apparent (projected) area. By creating different types of surfaces with a variety of input parameters, we create a database linking surface roughness parameters to the ratio. This database is used to train Machine Learning (ML) models and validate them appropriately. Specifically, we train deep, feed-forward neural networks and random forest models and validate them on a separate (held-out) test dataset. We investigate systematically the amount of input data needed to get accurate predictions on the test data. We also evaluate the importance of different input roughness parameters with respect to their effects on surface wettability. To this end, we study the weights that the learning AI models assigned to roughness parameters through training and discuss the findings with respect to experimental expectations.

Investigation of Machine Learning-based Coarse-Grained Mapping Schemes for Organic Molecules

Due to the wide range of timescales that are present in macromolecular systems, hierarchical mult... more Due to the wide range of timescales that are present in macromolecular systems, hierarchical multiscale strategies are necessary for their computational study. Coarse-graining (CG) allows to establish a link between different system resolutions and provides the backbone for the development of robust multiscale simulations and analyses. The CG mapping process is typically system-and application-specific, and it relies on chemical intuition. In this work, we explored the application of a Machine Learning strategy, based on Variational Autoencoders, for the development of suitable mapping schemes from the atomistic to the coarse-grained space of molecules with increasing chemical complexity. An extensive evaluation of the effect of the model hyperparameters on the training process and on the final output was performed, and an existing method was extended with the definition of different loss functions and the implementation of a selection criterion that ensures physical consistency of the output. The relationship between the input feature choice and the reconstruction accuracy was analyzed, supporting the need to introduce rotational invariance into the system. Strengths and limitations of the approach, both in the mapping and in the backmapping steps, are highlighted and critically discussed.

Developing Machine-Learned Potentials for Coarse-Grained Molecular Simulations: Challenges and Pitfalls

Coarse graining (CG) enables the investigation of molecular properties for larger systems and at ... more Coarse graining (CG) enables the investigation of molecular properties for larger systems and at longer timescales than the ones attainable at the atomistic resolution. Machine learning techniques have been recently proposed to learn CG particle interactions, i.e. develop CG force fields. Graph representations of molecules and supervised training of a graph convolutional neural network architecture are used to learn the potential of mean force through a force matching scheme. In this work, the force acting on each CG particle is correlated to a learned representation of its local environment that goes under the name of SchNet, constructed via continuous filter convolutions. We explore the application of SchNet models to obtain a CG potential for liquid benzene, investigating the effect of model architecture and hyperparameters on the thermodynamic, dynamical, and structural properties of the simulated CG systems, reporting and discussing challenges encountered and future directions envisioned. CCS CONCEPTS • Applied computing → Physical sciences and engineering • Computing methodologies → Machine learning → Machine learning approaches → Neural Networks • Computing methodologies → Modeling and simulation → Simulation types and techniques → Molecular simulation

Graph vs. bag representation models for the topic classification of web documents

World Wide Web, Aug 12, 2015

Text classification constitutes a popular task in Web research with various applications that ran... more Text classification constitutes a popular task in Web research with various applications that range from spam filtering to sentiment analysis. In this paper, we argue that its performance depends on the quality of Web documents, which varies significantly. For example, the curated content of news articles involves different challenges than the user-generated content of blog posts and Social Media messages. We experimentally verify our claim, quantifying the main factors that affect the performance of text classification. We also argue that the established bagof-words representation models are inadequate for handling all document types, as they merely extract frequent, yet distinguishing terms from the textual content of the training set. Thus, they suffer from low robustness in the context of noisy or unseen content, unless they are enriched with contextual, application-specific information. In their place, we propose the use of n-gram graphs, a model that goes beyond the bag-of-words representation, transforming every document into a graph: its nodes correspond to character or word n-grams and the co-occurring ones are connected by weighted edges. Individual document graphs can be combined into class graphs and graph similarities are employed to position and classify documents into the vector space. This approach offers two advantages with respect to bag models: first, classification accuracy increases due to the contextual information that is encapsulated in the edges of the n-gram graphs. Second, it reduces the search space to a limited set of robust, endogenous features that depend on the number of classes, rather than the size of the vocabulary. Our thorough experimental study over three large, real-world corpora confirms the superior performance of n-gram graphs across the main types of Web documents.

What Can Machine Learning Approaches in Genomics Tell Us about the Molecular Basis of Amyotrophic Lateral Sclerosis?

Journal of Personalized Medicine, Nov 26, 2020

Amyotrophic Lateral Sclerosis (ALS) is the most common late-onset motor neuron disorder, but our ... more Amyotrophic Lateral Sclerosis (ALS) is the most common late-onset motor neuron disorder, but our current knowledge of the molecular mechanisms and pathways underlying this disease remain elusive. This review (1) systematically identifies machine learning studies aimed at the understanding of the genetic architecture of ALS, (2) outlines the main challenges faced and compares the different approaches that have been used to confront them, and (3) compares the experimental designs and results produced by those approaches and describes their reproducibility in terms of biological results and the performances of the machine learning models. The majority of the collected studies incorporated prior knowledge of ALS into their feature selection approaches, and trained their machine learning models using genomic data combined with other types of mined knowledge including functional associations, protein-protein interactions, disease/tissue-specific information, epigenetic data, and known ALS phenotype-genotype associations. The importance of incorporating gene-gene interactions and cis-regulatory elements into the experimental design of future ALS machine learning studies is highlighted. Lastly, it is suggested that future advances in the genomic and machine learning fields will bring about a better understanding of ALS genetic architecture, and enable improved personalized approaches to this and other devastating and complex diseases.

Artificial Intelligence in the Deployment of Complex tasks in Pharmacology. A case of “Big Data” challenges in Ethnopharmacology

Planta Medica, Dec 1, 2021

Inverse Design of ZIFs through Artificial Intelligence Methods

Artificial Intelligence (AI) benefits research on membrane separations by facilitating fast and a... more Artificial Intelligence (AI) benefits research on membrane separations by facilitating fast and accurate performance predictions of a given material. However, the potential of AI to work backwards, towards predicting/designing a finetuned material for a given separation, remains untapped. Recent works report the inverse design of functionalized materials, such as metalorganic frameworks (MOFs), but they are limited to targeted sorption properties, while diffusivity, D, which is the driving force in membrane-based separations, is omitted. Herein, we report a tool combining a biologically inspired evolutionary algorithm with machine learning to design finetuned Zeolitic-Imidazolate Frameworks (ZIFs), a sub-family of MOFs, for desired sets of Di and Di/Dj of any given mixture of species i and j. We display the efficacy of our tool, by designing ZIFs that meet industrial performance criteria of permeability and selectivity, for CO2/CH4, O2/N2 and C3H6/C3H8 mixtures. We validate the designed ZIFs through appropriate simulations, confirming the suitability of the AI-suggested ZIF designs. Metal-organic frameworks (MOFs) [1] are multifunctional crystalline nanoporous media, which are studied for several diverse applications among which significant effort has been placed on the development of membranes for gas separations. However, modifying them towards an enhanced

The Entity Name System: Enabling the web of entities

We are currently witnessing an increasing interest in the use of the web as an information and kn... more We are currently witnessing an increasing interest in the use of the web as an information and knowledge source. Much of the information sought after in the web is in this case relevant to named entities (i.e., persons, locations, organizations, etc.). An important observation is that the entity identification problem lies at the core of many applications in this context. In order to deal with this problem, we propose the Entity Name System (ENS), a large scale, distributed infrastructure for assigning and managing unique identifiers for entities in the web. In this paper, we examine the special requirements for storage and management of entities, in the context of the ENS. We present a conceptual model for the representation of entities, and discuss problems related to data quality, as well as the management of the entity lifecycle. Finally, we describe the architecture of the current prototype of the system. 1 An example of a system that provides such functionality is SIGMA (http://www.sig.ma)

JedAI: The Force Behind Entity Resolution

Lecture Notes in Computer Science, 2017

We present JedAI, a toolkit for Entity Resolution that can be used in three different ways: as an... more We present JedAI, a toolkit for Entity Resolution that can be used in three different ways: as an open-source Java library that implements numerous state-of-the-art, domain-independent methods, as a workbench that facilitates the evaluation of their relative performance and as a desktop application that offers out-of-the-box ER solutions. JedAI bridges the gap between the database and the Semantic Web communities, offering solutions that are applicable to both relational and RDF data. It also conveys a modular architecture that facilitates its extension with more methods and with more comprehensive workflows.

Machine Learning Guided Atom Mapping of Metabolic Reactions

Journal of Chemical Information and Modeling, Nov 30, 2018

The atom mapping of a chemical reaction is a bijection between the atoms in the reactant molecule... more The atom mapping of a chemical reaction is a bijection between the atoms in the reactant molecules and the atoms in the product molecules. It encodes the underlying reaction mechanism and, as such, constitutes essential information in computational studies in metabolic engineering. Various techniques have been investigated for the automatic computation of the atom mapping of a chemical reaction approaching the problem as a graph matching problem. The graph abstraction of the chemical problem though, eliminates crucial chemical information. There have been efforts for enhancing the graph representation by introducing the bond stabilities as edge weights, as they are estimated based on experimental evidence. Here, we present a fully automated optimization-based approach, named AMLGAM (Automated Machine Learning Guided Atom Mapping), that uses machine learning techniques for the estimation of the bond stabilities based on the chemical environment of each bond. The optimization method finds the reaction mechanism which favors the breakage/formation of the less stable bonds. We evaluate our method on a manually curated dataset of 382 chemical reactions and we run our method on a much larger and diverse dataset of 7,400 chemical reactions. We show that the proposed method improves the accuracy over existing techniques, based on results published by earlier studies on a common dataset, and is capable of handling unbalanced reactions.

JedAI 3 : beyond batch, blocking-based Entity Resolution

Extending Database Technology, 2020

JedAI is an open-source toolkit that allows for building and benchmarking thousands of schema-agn... more JedAI is an open-source toolkit that allows for building and benchmarking thousands of schema-agnostic Entity Resolution (ER) pipelines through a non-learning, blocking-based end-to-end workflow. In this paper, we present its latest release, JedAI 3 , which conveys two new end-to-end workflows: one for budgetagnostic ER that is based on similarity joins, and one for budgetaware (i.e., progressive) ER. This version also adds support for pre-trained word or character embeddings and connects JedAI to the Python data analysis ecosystem. Overall, these enhancements provide JedAI with features offered by no other ER tool, especially in the schema-and domain-agnostic context.

Generating Realistic Nanorough Surfaces Using an N-Gram-Graph Augmented Deep Convolutional Generative Adversarial Network

Data mining for predicting gas diffusivity in zeolitic-imidazolate frameworks (ZIFs)

Journal of materials chemistry. A, Materials for energy and sustainability, 2022

Please note that technical editing may introduce minor changes to the text and/or graphics, which... more Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal's standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.

Summary Evaluation: Together We Stand NPowER-ed

Lecture Notes in Computer Science, 2013

ABSTRACT Summary evaluation has been a distinct domain of research for several years. Human summa... more ABSTRACT Summary evaluation has been a distinct domain of research for several years. Human summary evaluation appears to be a high-level cognitive process and, thus, difficult to reproduce. Even though several automatic evaluation methods correlate well to human evaluations over systems, we fail to get equivalent results when judging individual summaries. In this work, we propose the NPowER evaluation method based on machine learning and a set of methods from the family of &quot;n-gram graph&quot;-based summary evaluation methods. First, we show that the combined, optimized use of the evaluation methods outperforms the individual ones. Second, we compare the proposed method to a combination of ROUGE metrics. Third, we study and discuss what can make future evaluation measures better, based on the results of feature selection. We show that we can easily provide per summary evaluations that are far superior to existing performance of evaluation systems and face different measures under a unified view.

Reproducible experiments on Three-Dimensional Entity Resolution with JedAI

Information Systems, Dec 1, 2021

In Papadakis et al. [1], we presented the latest release of JedAI, an open-source Entity Resoluti... more In Papadakis et al. [1], we presented the latest release of JedAI, an open-source Entity Resolution (ER) system that allows for building a large variety of end-to-end ER pipelines. Through a thorough experimental evaluation, we compared a schema-agnostic ER pipeline based on blocks with another schema-based ER pipeline based on similarity joins. We applied them to 10 established, real-world datasets and assessed them with respect to effectiveness and time efficiency. Special care was taken to juxtapose their scalability, too, using seven established, synthetic datasets. Moreover, we experimentally compared the effectiveness of the batch schema-agnostic ER pipeline with its progressive counterpart. In this companion paper, we describe how to reproduce the entire experimental study that pertains to JedAI's serial execution through its intuitive user interface. We also explain how to examine the robustness of the parameter configurations we have selected.

GeoSensor is a novel, open-source system that enriches change detection over satellite images wit... more GeoSensor is a novel, open-source system that enriches change detection over satellite images with event detection over news items and social media content. GeoSensor combines these two orthogonal operations through state-of-the-art Semantic Web technologies. At its core lies the open-source, semantics-enabled Big Data infrastructure developed by the EU H2020 BigDataEurope project. This allows GeoSensor to offer an on-line functionality, despite facing three major challenges of Big Data: Volume (a single satellite image typically occupies a few GBs), Variety (its data sources include two different types of satellite images and various types of user-generated content) and Veracity, as the accuracy of the end result is crucial for the usefulness of our system. We present GeoSensor&amp;amp;amp;#39;s architecture in detail, highlighting the advantages of using semantics for taking the most of the knowledge extracted from news items and Earth Observation products. We also verify GeoSensor&amp;amp;amp;#39;s efficiency through a preliminary experimental study.

A Graph-based Text Similarity Measure That Employs Named Entity Information

Text comparison is an interesting though hard task, with many applications in Natural Language Pr... more Text comparison is an interesting though hard task, with many applications in Natural Language Processing. This work introduces a new text-similarity measure, which employs named-entities' information extracted from the texts and the ngram graphs' model for representing documents. Using OpenCalais as a namedentity recognition service and the JIN-SECT toolkit for constructing and managing n-gram graphs, the text similarity measure is embedded in a text clustering algorithm (k-Means). The evaluation of the produced clusters with various clustering validity metrics shows that the extraction of named entities at a first step can be profitable for the time-performance of similarity measures that are based on the n-gram graph representation without affecting the overall performance of the NLP task.

Detecting and exploiting stability in evolving heterogeneous information spaces

Individuals contribute content on the Web at an unprecedented rate, accumulating immense quantiti... more Individuals contribute content on the Web at an unprecedented rate, accumulating immense quantities of (semi-) structured data. Wisdom of the Crowds theory advocates that such information (or parts of it) is constantly overwritten, updated, or even deleted by other users, with the goal of rendering it more accurate, or up-to-date. This is particularly true for the collaboratively edited, semi-structured data of entity repositories, whose entity profiles are consistently kept fresh. Therefore, their core information that remain stable with the passage of time, despite being reviewed by numerous users, are particularly useful for the description of an entity. Based on the above hypothesis, we introduce a classification scheme that predicts, on the basis of statistical and content patterns, whether an attribute (i.e., name-value pair) is going to be modified in the future. We apply our scheme on a large, real-world, versioned dataset and verify its effectiveness. Our thorough experimental study also suggests that reducing entity profiles to their stable parts conveys significant benefits to two common tasks in computer science: information retrieval and information integration.

Representation models for text classification

Text classification constitutes a popular task in Web research with various applications that ran... more Text classification constitutes a popular task in Web research with various applications that range from spam filtering to sentiment analysis. To address it, patterns of cooccurring words or characters are typically extracted from the textual content of Web documents. However, not all documents are of the same quality; for example, the curated content of news articles usually entails lower levels of noise than the user-generated content of the blog posts and the other Social Media. In this paper, we provide some insight and a preliminary study on a tripartite categorization of Web documents, based on inherent document characteristics. We claim and support that each category calls for different classification settings with respect to the representation model. We verify this claim experimentally, by showing that topic classification on these different document types offers very different results per type. In addition, we consider a novel approach that improves the performance of topic classification across all types of Web documents: namely the n-gram graphs. This model goes beyond the established bag-of-words one, representing each document as a graph. Individual graphs can be combined into a class graph and graph similarities are then employed to position and classify documents into the vector space. Accuracy is increased due to the contextual information that is encapsulated in the edges of the n-gram graphs; efficiency, on the other hand, is boosted by reducing the feature space to a limited set of dimensions that depend on the number of classes, rather than the size of the vocabulary. Our experimental study over three largescale, real-world data sets validates the higher performance of n-gram graphs in all three domains of Web documents.

Three-dimensional Entity Resolution with JedAI

Information Systems, Nov 1, 2020

Abstract Entity Resolution (ER) is the task of detecting different entity profiles that describe ... more Abstract Entity Resolution (ER) is the task of detecting different entity profiles that describe the same real-world objects. To facilitate its execution, we have developed JedAI , an open-source system that puts together a series of state-of-the-art ER techniques that have been proposed and examined independently, targeting parts of the ER end-to-end pipeline. This is a unique approach, as no other ER tool brings together so many established techniques. Instead, most ER tools merely convey a few techniques, those primarily developed by their creators. In addition to democratizing ER techniques, JedAI goes beyond the other ER tools by offering a series of unique characteristics: (i) It allows for building and benchmarking millions of ER pipelines. (ii) It is the only ER system that applies seamlessly to any combination of structured and/or semi-structured data. (iii) It constitutes the only ER system that runs seamlessly both on stand-alone computers and clusters of computers — through the parallel implementation of all algorithms in Apache Spark. (iv) It supports two different end-to-end workflows for carrying out batch ER (i.e., budget-agnostic), a schema-agnostic one based on blocks, and a schema-based one relying on similarity joins. (v) It adapts both end-to-end workflows to budget-aware (i.e., progressive) ER. We present in detail all features of JedAI , stressing the core characteristics that enhance its usability, and boost its versatility and effectiveness. We also compare it to the state-of-the-art in the field, qualitatively and quantitatively, demonstrating its state-of-the-art performance over a variety of large-scale datasets from different domains. The central repository of the JedAI ’s code base is here: https://github.com/scify/JedAIToolkit . A video demonstrating the JedAI’s Web application is available here: https://www.youtube.com/watch?v=OJY1DUrUAe8 .

Hybridizing AI and Domain Knowledge in Nanotechnology: the Example of Surface Roughness Effects on Wetting Behavior

In this paper, we propose a scheme for the hybridization of domain modeling and theoretical knowl... more In this paper, we propose a scheme for the hybridization of domain modeling and theoretical knowledge in nanotechnology with Artificial Intelligence(AI) techniques and evaluate the success of its application to predict the relationship between nanosurface morphology and wettability. We utilize domain knowledge consisting of two parts. The first part is a mathematical modeling based on the inverse Fourier transform for the generation of rough surfaces with Gaussian or non-Gaussian height distributions, characterized by their first moments (Rms, skewness, kurtosis) and the correlation lengths along x and y-axes. The second part lies in the assumption that the Wenzel scenario for wetting of rough surfaces holds where the critical parameter for contact angle determination is the roughness ratio , defined as the ratio of true (active) area of the solid surface to the apparent (projected) area. By creating different types of surfaces with a variety of input parameters, we create a database linking surface roughness parameters to the ratio. This database is used to train Machine Learning (ML) models and validate them appropriately. Specifically, we train deep, feed-forward neural networks and random forest models and validate them on a separate (held-out) test dataset. We investigate systematically the amount of input data needed to get accurate predictions on the test data. We also evaluate the importance of different input roughness parameters with respect to their effects on surface wettability. To this end, we study the weights that the learning AI models assigned to roughness parameters through training and discuss the findings with respect to experimental expectations.

Investigation of Machine Learning-based Coarse-Grained Mapping Schemes for Organic Molecules

Due to the wide range of timescales that are present in macromolecular systems, hierarchical mult... more Due to the wide range of timescales that are present in macromolecular systems, hierarchical multiscale strategies are necessary for their computational study. Coarse-graining (CG) allows to establish a link between different system resolutions and provides the backbone for the development of robust multiscale simulations and analyses. The CG mapping process is typically system-and application-specific, and it relies on chemical intuition. In this work, we explored the application of a Machine Learning strategy, based on Variational Autoencoders, for the development of suitable mapping schemes from the atomistic to the coarse-grained space of molecules with increasing chemical complexity. An extensive evaluation of the effect of the model hyperparameters on the training process and on the final output was performed, and an existing method was extended with the definition of different loss functions and the implementation of a selection criterion that ensures physical consistency of the output. The relationship between the input feature choice and the reconstruction accuracy was analyzed, supporting the need to introduce rotational invariance into the system. Strengths and limitations of the approach, both in the mapping and in the backmapping steps, are highlighted and critically discussed.

Developing Machine-Learned Potentials for Coarse-Grained Molecular Simulations: Challenges and Pitfalls

Coarse graining (CG) enables the investigation of molecular properties for larger systems and at ... more Coarse graining (CG) enables the investigation of molecular properties for larger systems and at longer timescales than the ones attainable at the atomistic resolution. Machine learning techniques have been recently proposed to learn CG particle interactions, i.e. develop CG force fields. Graph representations of molecules and supervised training of a graph convolutional neural network architecture are used to learn the potential of mean force through a force matching scheme. In this work, the force acting on each CG particle is correlated to a learned representation of its local environment that goes under the name of SchNet, constructed via continuous filter convolutions. We explore the application of SchNet models to obtain a CG potential for liquid benzene, investigating the effect of model architecture and hyperparameters on the thermodynamic, dynamical, and structural properties of the simulated CG systems, reporting and discussing challenges encountered and future directions envisioned. CCS CONCEPTS • Applied computing → Physical sciences and engineering • Computing methodologies → Machine learning → Machine learning approaches → Neural Networks • Computing methodologies → Modeling and simulation → Simulation types and techniques → Molecular simulation

Graph vs. bag representation models for the topic classification of web documents

World Wide Web, Aug 12, 2015

Text classification constitutes a popular task in Web research with various applications that ran... more Text classification constitutes a popular task in Web research with various applications that range from spam filtering to sentiment analysis. In this paper, we argue that its performance depends on the quality of Web documents, which varies significantly. For example, the curated content of news articles involves different challenges than the user-generated content of blog posts and Social Media messages. We experimentally verify our claim, quantifying the main factors that affect the performance of text classification. We also argue that the established bagof-words representation models are inadequate for handling all document types, as they merely extract frequent, yet distinguishing terms from the textual content of the training set. Thus, they suffer from low robustness in the context of noisy or unseen content, unless they are enriched with contextual, application-specific information. In their place, we propose the use of n-gram graphs, a model that goes beyond the bag-of-words representation, transforming every document into a graph: its nodes correspond to character or word n-grams and the co-occurring ones are connected by weighted edges. Individual document graphs can be combined into class graphs and graph similarities are employed to position and classify documents into the vector space. This approach offers two advantages with respect to bag models: first, classification accuracy increases due to the contextual information that is encapsulated in the edges of the n-gram graphs. Second, it reduces the search space to a limited set of robust, endogenous features that depend on the number of classes, rather than the size of the vocabulary. Our thorough experimental study over three large, real-world corpora confirms the superior performance of n-gram graphs across the main types of Web documents.

What Can Machine Learning Approaches in Genomics Tell Us about the Molecular Basis of Amyotrophic Lateral Sclerosis?

Journal of Personalized Medicine, Nov 26, 2020

Amyotrophic Lateral Sclerosis (ALS) is the most common late-onset motor neuron disorder, but our ... more Amyotrophic Lateral Sclerosis (ALS) is the most common late-onset motor neuron disorder, but our current knowledge of the molecular mechanisms and pathways underlying this disease remain elusive. This review (1) systematically identifies machine learning studies aimed at the understanding of the genetic architecture of ALS, (2) outlines the main challenges faced and compares the different approaches that have been used to confront them, and (3) compares the experimental designs and results produced by those approaches and describes their reproducibility in terms of biological results and the performances of the machine learning models. The majority of the collected studies incorporated prior knowledge of ALS into their feature selection approaches, and trained their machine learning models using genomic data combined with other types of mined knowledge including functional associations, protein-protein interactions, disease/tissue-specific information, epigenetic data, and known ALS phenotype-genotype associations. The importance of incorporating gene-gene interactions and cis-regulatory elements into the experimental design of future ALS machine learning studies is highlighted. Lastly, it is suggested that future advances in the genomic and machine learning fields will bring about a better understanding of ALS genetic architecture, and enable improved personalized approaches to this and other devastating and complex diseases.

Artificial Intelligence in the Deployment of Complex tasks in Pharmacology. A case of “Big Data” challenges in Ethnopharmacology

Planta Medica, Dec 1, 2021

Inverse Design of ZIFs through Artificial Intelligence Methods

Artificial Intelligence (AI) benefits research on membrane separations by facilitating fast and a... more Artificial Intelligence (AI) benefits research on membrane separations by facilitating fast and accurate performance predictions of a given material. However, the potential of AI to work backwards, towards predicting/designing a finetuned material for a given separation, remains untapped. Recent works report the inverse design of functionalized materials, such as metalorganic frameworks (MOFs), but they are limited to targeted sorption properties, while diffusivity, D, which is the driving force in membrane-based separations, is omitted. Herein, we report a tool combining a biologically inspired evolutionary algorithm with machine learning to design finetuned Zeolitic-Imidazolate Frameworks (ZIFs), a sub-family of MOFs, for desired sets of Di and Di/Dj of any given mixture of species i and j. We display the efficacy of our tool, by designing ZIFs that meet industrial performance criteria of permeability and selectivity, for CO2/CH4, O2/N2 and C3H6/C3H8 mixtures. We validate the designed ZIFs through appropriate simulations, confirming the suitability of the AI-suggested ZIF designs. Metal-organic frameworks (MOFs) [1] are multifunctional crystalline nanoporous media, which are studied for several diverse applications among which significant effort has been placed on the development of membranes for gas separations. However, modifying them towards an enhanced

The Entity Name System: Enabling the web of entities

We are currently witnessing an increasing interest in the use of the web as an information and kn... more We are currently witnessing an increasing interest in the use of the web as an information and knowledge source. Much of the information sought after in the web is in this case relevant to named entities (i.e., persons, locations, organizations, etc.). An important observation is that the entity identification problem lies at the core of many applications in this context. In order to deal with this problem, we propose the Entity Name System (ENS), a large scale, distributed infrastructure for assigning and managing unique identifiers for entities in the web. In this paper, we examine the special requirements for storage and management of entities, in the context of the ENS. We present a conceptual model for the representation of entities, and discuss problems related to data quality, as well as the management of the entity lifecycle. Finally, we describe the architecture of the current prototype of the system. 1 An example of a system that provides such functionality is SIGMA (http://www.sig.ma)

JedAI: The Force Behind Entity Resolution

Lecture Notes in Computer Science, 2017

We present JedAI, a toolkit for Entity Resolution that can be used in three different ways: as an... more We present JedAI, a toolkit for Entity Resolution that can be used in three different ways: as an open-source Java library that implements numerous state-of-the-art, domain-independent methods, as a workbench that facilitates the evaluation of their relative performance and as a desktop application that offers out-of-the-box ER solutions. JedAI bridges the gap between the database and the Semantic Web communities, offering solutions that are applicable to both relational and RDF data. It also conveys a modular architecture that facilitates its extension with more methods and with more comprehensive workflows.

Machine Learning Guided Atom Mapping of Metabolic Reactions

Journal of Chemical Information and Modeling, Nov 30, 2018

The atom mapping of a chemical reaction is a bijection between the atoms in the reactant molecule... more The atom mapping of a chemical reaction is a bijection between the atoms in the reactant molecules and the atoms in the product molecules. It encodes the underlying reaction mechanism and, as such, constitutes essential information in computational studies in metabolic engineering. Various techniques have been investigated for the automatic computation of the atom mapping of a chemical reaction approaching the problem as a graph matching problem. The graph abstraction of the chemical problem though, eliminates crucial chemical information. There have been efforts for enhancing the graph representation by introducing the bond stabilities as edge weights, as they are estimated based on experimental evidence. Here, we present a fully automated optimization-based approach, named AMLGAM (Automated Machine Learning Guided Atom Mapping), that uses machine learning techniques for the estimation of the bond stabilities based on the chemical environment of each bond. The optimization method finds the reaction mechanism which favors the breakage/formation of the less stable bonds. We evaluate our method on a manually curated dataset of 382 chemical reactions and we run our method on a much larger and diverse dataset of 7,400 chemical reactions. We show that the proposed method improves the accuracy over existing techniques, based on results published by earlier studies on a common dataset, and is capable of handling unbalanced reactions.