Automatic Knowledge Base Construction Using Probabilistic Extraction, Deductive Reasoning, and Human Feedback
Total Page:16
File Type:pdf, Size:1020Kb
Load more
Recommended publications
-
Knowledge Extraction by Internet Monitoring to Enhance Crisis Management
El Hamali Knowledge Extraction by Internet Monitoring to Enhance Crisis Management Knowledge Extraction by Internet Monitoring to Enhance Crisis Management El Hamali Samiha Nadia Nouali-TAboudjnent, Omar Nouali National School for Computer Science ESI, Research Center of Scientific and Technique Algiers, Algeria Information CERIST, Algeria [email protected] {nnouali, onouali}@cerist.dz ABSTRACT This paper presents our work on developing a system for Internet monitoring and knowledge extraction from different web documents which contain information about disasters. This system is based on ontology of the disasters domain for the knowledge extraction and it presents all the information extracted according to the kind of the disaster defined in the ontology. The system disseminates the information extracted (as a synthesis of the web documents) to the users after a filtering based on their profiles. The profile of a user is updated automatically by interactively taking into account his feedback. Keywords Crisis management, internet monitoring, knowledge extraction, ontology, information filtering. INTRODUCTION Several crises and disasters happened in the last decades. After the southern Asian tsunami, the Katrina hurricane in the United States, and the Kashmir earthquake, people throughout the world used the Web as a source of information and a means of communication. In the hours and days immediately following a disaster, a lot of information becomes available on the web. With a single Google search, 17 millions “emergency management” sites and documents have been listed (Min & Peishih, 2008). The user finds a lot of information which comes from different sources, for instance : during the first 3 days of tsunami disaster in south-east Asia, 4000 reports were gathered by the Google News service, and within a month of the incident, 200000 distinct news reports from several online sources (Yiming, et al., 2007 IEEE). -
Injection of Automatically Selected Dbpedia Subjects in Electronic
Injection of Automatically Selected DBpedia Subjects in Electronic Medical Records to boost Hospitalization Prediction Raphaël Gazzotti, Catherine Faron Zucker, Fabien Gandon, Virginie Lacroix-Hugues, David Darmon To cite this version: Raphaël Gazzotti, Catherine Faron Zucker, Fabien Gandon, Virginie Lacroix-Hugues, David Darmon. Injection of Automatically Selected DBpedia Subjects in Electronic Medical Records to boost Hos- pitalization Prediction. SAC 2020 - 35th ACM/SIGAPP Symposium On Applied Computing, Mar 2020, Brno, Czech Republic. 10.1145/3341105.3373932. hal-02389918 HAL Id: hal-02389918 https://hal.archives-ouvertes.fr/hal-02389918 Submitted on 16 Dec 2019 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Injection of Automatically Selected DBpedia Subjects in Electronic Medical Records to boost Hospitalization Prediction Raphaël Gazzotti Catherine Faron-Zucker Fabien Gandon Université Côte d’Azur, Inria, CNRS, Université Côte d’Azur, Inria, CNRS, Inria, Université Côte d’Azur, CNRS, I3S, Sophia-Antipolis, France I3S, Sophia-Antipolis, France -
How to Keep a Knowledge Base Synchronized with Its Encyclopedia Source
Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence (IJCAI-17) How to Keep a Knowledge Base Synchronized with Its Encyclopedia Source Jiaqing Liang12, Sheng Zhang1, Yanghua Xiao134∗ 1School of Computer Science, Shanghai Key Laboratory of Data Science Fudan University, Shanghai, China 2Shuyan Technology, Shanghai, China 3Shanghai Internet Big Data Engineering Technology Research Center, China 4Xiaoi Research, Shanghai, China [email protected], fshengzhang16,[email protected] Abstract However, most of these knowledge bases tend to be out- dated, which limits their utility. For example, in many knowl- Knowledge bases are playing an increasingly im- edge bases, Donald Trump is only a business man even af- portant role in many real-world applications. How- ter the inauguration. Obviously, it is important to let ma- ever, most of these knowledge bases tend to be chines know that Donald Trump is the president of the United outdated, which limits the utility of these knowl- States so that they can understand that the topic of an arti- edge bases. In this paper, we investigate how to cle mentioning Donald Trump is probably related to politics. keep the freshness of the knowledge base by syn- Moreover, new entities are continuously emerging and most chronizing it with its data source (usually ency- of them are popular, such as iphone 8. However, it is hard for clopedia websites). A direct solution is revisiting a knowledge base to cover these entities in time even if the the whole encyclopedia periodically and rerun the encyclopedia websites have already covered them. entire pipeline of the construction of knowledge freshness base like most existing methods. -
Handling Semantic Complexity of Big Data Using Machine Learning and RDF Ontology Model
S S symmetry Article Handling Semantic Complexity of Big Data using Machine Learning and RDF Ontology Model Rauf Sajjad *, Imran Sarwar Bajwa and Rafaqut Kazmi Department of Computer Science & IT, The Islamia University Bahawalpur, Bahawalpur 63100, Pakistan; [email protected] (I.S.B.); [email protected] (R.K.) * Correspondence: [email protected]; Tel.: +92-62-925-5466 Received: 3 January 2019; Accepted: 14 February 2019; Published: 1 March 2019 Abstract: Business information required for applications and business processes is extracted using systems like business rule engines. Since the advent of Big Data, such rule engines are producing rules in a big quantity whereas more rules lead to more complexity in semantic analysis and understanding. This paper introduces a method to handle semantic complexity in rules and support automated generation of Resource Description Framework (RDF) metadata model of rules and such model is used to assist in querying and analysing Big Data. Practically, the dynamic changes in rules can be a source of conflict in rules stored in a repository. It is identified during the literature review that there is a need of a method that can semantically analyse rules and help business analysts in testing and validating the rules once a change is made in a rule. This paper presents a robust method that not only supports semantic analysis of rules but also generates RDF metadata model of rules and provide support of querying for the sake of semantic interpretation of the rules. The results of the experiments manifest that consistency checking of a set of big data rules is possible through automated tools. -
Learning Ontologies from RDF Annotations
/HDUQLQJÃRQWRORJLHVÃIURPÃ5')ÃDQQRWDWLRQV $OH[DQGUHÃ'HOWHLOÃ&DWKHULQHÃ)DURQ=XFNHUÃ5RVHÃ'LHQJ ACACIA project, INRIA, 2004, route des Lucioles, B.P. 93, 06902 Sophia Antipolis, France {Alexandre.Delteil, Catherine.Faron, Rose.Dieng}@sophia.inria.fr $EVWUDFW objects, as in [Mineau, 1990; Carpineto and Romano, 1993; Bournaud HWÃDO., 2000]. In this paper, we present a method for learning Since all RDF annotations are gathered inside a common ontologies from RDF annotations of Web RDF graph, the problem which arises is the extraction of a resources by systematically generating the most description for a given resource from the whole RDF graph. specific generalization of all the possible sets of After a brief description of the RDF data model (Section 2) resources. The preliminary step of our method and of RDF Schema (Section 3), Section 4 presents several consists in extracting (partial) resource criteria for extracting partial resource descriptions. In order descriptions from the whole RDF graph gathering to deal with the intrinsic complexity of the building of a all the annotations. In order to deal with generalization hierarchy, we propose an incremental algorithmic complexity, we incrementally build approach by gradually increasing the size of the descriptions the ontology by gradually increasing the size of the resource descriptions we consider. we consider. The principle of the approach is explained in Section 5 and more deeply detailed in Section 6. Ã ,QWURGXFWLRQ Ã 7KHÃ5')ÃGDWDÃPRGHO The Semantic Web, expected to be the next step that will RDF is the emerging Web standard for annotating resources lead the Web to its full potential, will be based on semantic HWÃDO metadata describing all kinds of Web resources. -
A Comparison of Knowledge Extraction Tools for the Semantic Web
A Comparison of Knowledge Extraction Tools for the Semantic Web Aldo Gangemi1;2 1 LIPN, Universit´eParis13-CNRS-SorbonneCit´e,France 2 STLab, ISTC-CNR, Rome, Italy. Abstract. In the last years, basic NLP tasks: NER, WSD, relation ex- traction, etc. have been configured for Semantic Web tasks including on- tology learning, linked data population, entity resolution, NL querying to linked data, etc. Some assessment of the state of art of existing Knowl- edge Extraction (KE) tools when applied to the Semantic Web is then desirable. In this paper we describe a landscape analysis of several tools, either conceived specifically for KE on the Semantic Web, or adaptable to it, or even acting as aggregators of extracted data from other tools. Our aim is to assess the currently available capabilities against a rich palette of ontology design constructs, focusing specifically on the actual semantic reusability of KE output. 1 Introduction We present a landscape analysis of the current tools for Knowledge Extraction from text (KE), when applied on the Semantic Web (SW). Knowledge Extraction from text has become a key semantic technology, and has become key to the Semantic Web as well (see. e.g. [31]). Indeed, interest in ontology learning is not new (see e.g. [23], which dates back to 2001, and [10]), and an advanced tool like Text2Onto [11] was set up already in 2005. However, interest in KE was initially limited in the SW community, which preferred to concentrate on manual design of ontologies as a seal of quality. Things started changing after the linked data bootstrapping provided by DB- pedia [22], and the consequent need for substantial population of knowledge bases, schema induction from data, natural language access to structured data, and in general all applications that make joint exploitation of structured and unstructured content. -
CN-Dbpedia: a Never-Ending Chinese Knowledge Extraction System
CN-DBpedia: A Never-Ending Chinese Knowledge Extraction System Bo Xu1,YongXu1, Jiaqing Liang1,2, Chenhao Xie1,2, Bin Liang1, B Wanyun Cui1, and Yanghua Xiao1,3( ) 1 Shanghai Key Laboratory of Data Science, School of Computer Science, Fudan University, Shanghai, China {xubo,yongxu16,jqliang15,xiech15,liangbin,shawyh}@fudan.edu.cn, [email protected] 2 Data Eyes Research, Shanghai, China 3 Shanghai Internet Big Data Engineering and Technology Center, Shanghai, China Abstract. Great efforts have been dedicated to harvesting knowledge bases from online encyclopedias. These knowledge bases play impor- tant roles in enabling machines to understand texts. However, most cur- rent knowledge bases are in English and non-English knowledge bases, especially Chinese ones, are still very rare. Many previous systems that extract knowledge from online encyclopedias, although are applicable for building a Chinese knowledge base, still suffer from two challenges. The first is that it requires great human efforts to construct an ontology and build a supervised knowledge extraction model. The second is that the update frequency of knowledge bases is very slow. To solve these chal- lenges, we propose a never-ending Chinese Knowledge extraction system, CN-DBpedia, which can automatically generate a knowledge base that is of ever-increasing in size and constantly updated. Specially, we reduce the human costs by reusing the ontology of existing knowledge bases and building an end-to-end facts extraction model. We further propose a smart active update strategy to keep the freshness of our knowledge base with little human costs. The 164 million API calls of the published services justify the success of our system. -
Knowledge Extraction for Hybrid Question Answering
KNOWLEDGEEXTRACTIONFORHYBRID QUESTIONANSWERING Von der Fakultät für Mathematik und Informatik der Universität Leipzig angenommene DISSERTATION zur Erlangung des akademischen Grades Doctor rerum naturalium (Dr. rer. nat.) im Fachgebiet Informatik vorgelegt von Ricardo Usbeck, M.Sc. geboren am 01.04.1988 in Halle (Saale), Deutschland Die Annahme der Dissertation wurde empfohlen von: 1. Professor Dr. Klaus-Peter Fähnrich (Leipzig) 2. Professor Dr. Philipp Cimiano (Bielefeld) Die Verleihung des akademischen Grades erfolgt mit Bestehen der Verteidigung am 17. Mai 2017 mit dem Gesamtprädikat magna cum laude. Leipzig, den 17. Mai 2017 bibliographic data title: Knowledge Extraction for Hybrid Question Answering author: Ricardo Usbeck statistical information: 10 chapters, 169 pages, 28 figures, 32 tables, 8 listings, 5 algorithms, 178 literature references, 1 appendix part supervisors: Prof. Dr.-Ing. habil. Klaus-Peter Fähnrich Dr. Axel-Cyrille Ngonga Ngomo institution: Leipzig University, Faculty for Mathematics and Computer Science time frame: January 2013 - March 2016 ABSTRACT Over the last decades, several billion Web pages have been made available on the Web. The growing amount of Web data provides the world’s largest collection of knowledge.1 Most of this full-text data like blogs, news or encyclopaedic informa- tion is textual in nature. However, the increasing amount of structured respectively semantic data2 available on the Web fosters new search paradigms. These novel paradigms ease the development of natural language interfaces which enable end- users to easily access and benefit from large amounts of data without the need to understand the underlying structures or algorithms. Building a natural language Question Answering (QA) system over heteroge- neous, Web-based knowledge sources requires various building blocks. -
Real-Time Population of Knowledge Bases: Opportunities and Challenges
Real-time Population of Knowledge Bases: Opportunities and Challenges Ndapandula Nakashole, Gerhard Weikum Max Planck Institute for Informatics Saarbrucken,¨ Germany {nnakasho,weikum}@mpi-inf.mpg.de Abstract 2011; Nakashole 2011) which is now three years old. The NELL system (Carlson 2010) follows a Dynamic content is a frequently accessed part “never-ending” extraction model with the extraction of the Web. However, most information ex- process going on 24 hours a day. However NELL’s traction approaches are batch-oriented, thus focus is on language learning by iterating mostly not effective for gathering rapidly changing data. This paper proposes a model for fact on the same ClueWeb09 corpus. Our focus is on extraction in real-time. Our model addresses capturing the latest information enriching it into the the difficult challenges that timely fact extrac- form of relational facts. Web-based news aggrega- tion on frequently updated data entails. We tors such as Google News and Yahoo! News present point out a naive solution to the main research up-to-date information from various news sources. question and justify the choices we make in However, news aggregators present headlines and the model we propose. short text snippets. Our focus is on presenting this information as relational facts that can facilitate re- 1 Introduction lational queries spanning new and historical data. Challenges. Timely knowledge extraction from fre- Motivation. Dynamic content is an important part quently updated sources entails a number of chal- of the Web, it accounts for a substantial amount of lenges: Web traffic. For example, much time is spent read- ing news, blogs and user comments in social media. -
Domain-Targeted, High Precision Knowledge Extraction
Domain-Targeted, High Precision Knowledge Extraction Bhavana Dalvi Mishra Niket Tandon Peter Clark Allen Institute for Artificial Intelligence 2157 N Northlake Way Suite 110, Seattle, WA 98103 bhavanad,nikett,peterc @allenai.org { } Abstract remains elusive. Specifically, our goal is a large, high precision body of (subject,predicate,object) Our goal is to construct a domain-targeted, statements relevant to elementary science, to sup- high precision knowledge base (KB), contain- port a downstream QA application task. Although ing general (subject,predicate,object) state- ments about the world, in support of a down- there are several impressive, existing resources that stream question-answering (QA) application. can contribute to our endeavor, e.g., NELL (Carlson Despite recent advances in information extrac- et al., 2010), ConceptNet (Speer and Havasi, 2013), tion (IE) techniques, no suitable resource for WordNet (Fellbaum, 1998), WebChild (Tandon et our task already exists; existing resources are al., 2014), Yago (Suchanek et al., 2007), FreeBase either too noisy, too named-entity centric, or (Bollacker et al., 2008), and ReVerb-15M (Fader et too incomplete, and typically have not been al., 2011), their applicability is limited by both constructed with a clear scope or purpose. limited coverage of general knowledge (e.g., To address these, we have created a domain- • targeted, high precision knowledge extraction FreeBase and NELL primarily contain knowl- pipeline, leveraging Open IE, crowdsourcing, edge about Named Entities; WordNet uses only and a novel canonical schema learning algo- a few (< 10) semantic relations) rithm (called CASI), that produces high pre- low precision (e.g., many ConceptNet asser- • cision knowledge targeted to a particular do- tions express idiosyncratic rather than general main - in our case, elementary science. -
KBART: Knowledge Bases and Related Tools
NISO-RP-9-2010 KBART: Knowledge Bases and Related Tools A Recommended Practice of the National Information Standards Organization (NISO) and UKSG Prepared by the NISO/UKSG KBART Working Group January 2010 i About NISO Recommended Practices A NISO Recommended Practice is a recommended "best practice" or "guideline" for methods, materials, or practices in order to give guidance to the user. Such documents usually represent a leading edge, exceptional model, or proven industry practice. All elements of Recommended Practices are discretionary and may be used as stated or modified by the user to meet specific needs. This recommended practice may be revised or withdrawn at any time. For current information on the status of this publication contact the NISO office or visit the NISO website (www.niso.org). Published by National Information Standards Organization (NISO) One North Charles Street, Suite 1905 Baltimore, MD 21201 www.niso.org Copyright © 2010 by the National Information Standards Organization and the UKSG. All rights reserved under International and Pan-American Copyright Conventions. For noncommercial purposes only, this publication may be reproduced or transmitted in any form or by any means without prior permission in writing from the publisher, provided it is reproduced accurately, the source of the material is identified, and the NISO/UKSG copyright status is acknowledged. All inquires regarding translations into other languages or commercial reproduction or distribution should be addressed to: NISO, One North Charles Street, -
4Th INCF Japan Node International Workshop Advances in Neuroinformatics 2016 and 14Th INCF Nodes Workshop
4th INCF Japan Node International Workshop Advances in Neuroinformatics 2016 and 14th INCF Nodes Workshop Program and Abstracts May 28-29, 2016 Suzuki Umetaro Hall, RIKEN 2-1, Hirosawa, Wako, Saitama, Japan RIKEN Brain Science Institute JAPAN Organized by RIKEN Brain Science Institute International Neuroinformatics Coordinating Facility (INCF) In Cooperation with The Japanese Society for Artificial Intelligence Japanese Neural Network Society The Japan Neuroscience Society ※This Workshop is a part of the RIKEN Symposium Series. AINI2016 & INCF Nodes Workshop Table of Contents Message from Chairs ..................................................................................................... 3 Workshop Organizing Committee ........................................................................... 4 General Information ....................................................................................................... 5 Program at a Glance ....................................................................................................... 6 Program ............................................................................................................................... 7 Keynote Lecture Abstracts ....................................................................................... 12 Oral Session I Whole Brain Imaging for Approaches to Functions and Anatomy Abstracts ...................................................................................................... 15 Oral Session II Multi-dimensional Theoretical Neuroscience