Scalable and Declarative Information Extraction in a Parallel Data Analytics System

Total Page:16

File Type:pdf, Size:1020Kb

Scalable and Declarative Information Extraction in a Parallel Data Analytics System Scalable and Declarative Information Extraction in a Parallel Data Analytics System DISSERTATION zur Erlangung des akademischen Grades Doktor-Ingenieur (Dr.-Ing.) im Fach Informatik eingereicht an der Mathematisch-Naturwissenschaftlichen Fakultät Humboldt-Universität zu Berlin von Dipl.-Inf. Astrid Rheinländer Präsidentin der Humboldt-Universität zu Berlin: Prof. Dr.-Ing. habil. Dr. Sabine Kunst Dekan der Mathematisch-Naturwissenschaftlichen Fakultät: Prof. Dr. Elmar Kulke Gutachter: 1. Prof. Dr.-Ing. Ulf Leser 2. Prof. Dr. Felix Naumann 3. Prof. Dr.-Ing. Norbert Ritter eingereicht am: 24.01.2017 Tag der mündlichen Prüfung: 23.06.2017 Zusammenfassung Die Menge unstrukturierter Daten ist in den letzten Jahren enorm gewachsen und in diesem Zusammenhang hat sich auch die Analysekomplexität solcher Daten wesentlich erhöht. Informationsextraktion (IE) ist ein bedeutendes Verfahren für viele Anwendun- gen, in denen unstrukturierte Texte in strukturierte Daten transformiert werden, jedoch erfordert die systematische Anwendung von IE-Techniken auf sehr große Datenmengen hochkomplexe, skalierbare und anpassungsfähige Systeme. Obwohl bereits eine umfan- greiche Sammlung von IE-Werkzeugen und Algorithmen für verschiedene IE-Aufgaben existiert, ist die nahtlose und erweiterbare Kombination dieser Werkzeuge in einem skalierbaren end-to-end IE-System immer noch eine große Herausforderung. Diese Dissertation untersucht genau diese Problemstellung, d.h., es wird ein anfrage- basiertes IE-System innerhalb einer parallelen Datenanalyseplattform erforscht und en- twickelt, das für konkrete Anwendungsdomänen konfigurierbar ist und für Textsamm- lungen im Terabyte-Bereich skaliert. Innerhalb dieses Forschungsfeldes werden vier konsekutive Forschungsfragen bearbeitet. Zuerst werden konfigurierbare, algebra- ische Operatoren für alle grundlegenden IE-Aufgaben und für Web Text Analytics (WA) definiert. Es wird gezeigt wie diese Operatoren genutzt werden können um kom- plexe IE-Aufgaben in Form von Queries innerhalb der deklarativen Anfragesprache Meteor auszudrücken. Solche Queries werden in algebraische Data Flows übersetzt, analysiert, logisch und physikalisch optimiert und schließlich in parallele Data Flow- Programme übersetzt, die mit der parallelen Datenanalyseplattform Stratosphere aus- geführt werden. Alle Operatoren werden hinsichtlich ihrer physikalischen, algebrais- chen und Laufzeiteigenschaften charakterisiert um sowohl das Potenzial als auch die Bedeutung der Optimierung der Ausführungsreihenfolge nicht-relationaler, benutzer- definierter Operatoren für Data Flows (UDFs) hervorzuheben. Als zweite Forschungs- frage wird der Stand der Technik in der Optimierung nicht-relationaler Data Flows untersucht. Relevante Optimierungstechniken, die in verschiedenen Phasen des Op- timierungsprozesses in parallelen Datenanalysesystemen eingesetzt werden, werden vorgestellt und existierende Data Flow-Anfragesprachen werden umfassend hinsichtlich der verfügbaren Optimierungstechniken analysiert. Die Analyse kommt zu dem Schluss, dass eine umfassende Optimierung von UDFs für viele Systeme immer noch eine Her- ausforderung ist. Basierend auf dieser Beobachtung schließt sich die dritte Forschungs- frage an, in der ein erweiterbarer, logischer Optimierer erforscht und entwickelt wird, der die Semantik von UDFs mit in den Optimierungsprozess mit einbezieht (SOFA). SOFA analysiert eine kompakte Menge von Eigenschaften, die die Semantik der UDFs beschreiben und kombiniert die automatisierte Analyse mit manuellen UDF-Annotatio- nen, um eine umfassende Optimierung von Data Flows zu ermöglichen. SOFA ist in der Lage, beliebige Data Flows aus unterschiedlichen Anwendungsbereichen logisch zu op- timieren, was zu erheblichen Laufzeitverbesserungen im Vergleich mit anderen Tech- niken führt. Als Viertes wird die Anwendbarkeit des vorgestellten IE-Systems auf real- weltliche Textsammlungen im Terabyte-Bereich untersucht, in dem Inhalte des World Wide Webs zu gesundheitsrelevanten Themen mit wissenschaftlichen Veröffentlichun- gen verglichen werden. Im Rahmen dieser Studie wird systematisch die Skalierbarkeit und Robustheit der eingesetzten Methoden und Werkzeuge untersucht sowie die Qual- ität der extrahierten Daten analysiert um schließlich die kritischsten Herausforderun- gen beim Aufbau eines IE-Systems für sehr große Datenmenge zu charakterisieren. iii Abstract In recent years, the size of unstructured data has grown tremendously and the com- plexity of the analysis of such data has increased significantly. In many domains, infor- mation extraction (IE) is an important technique to turn unstructured texts into struc- tured fact databases, but systematically applying IE techniques to very large inputs requires highly complex, adaptable, and scalable systems. Although a number of tools for different IE tasks exist, their seamless, extensible, and scalable combination into a large-scale end-to-end text analytics system still is a true challenge. This thesis addresses exactly this problem, i.e., we research and develop a query- based IE system that is accurate, configurable towards concrete application domains, and scalable to Terabyte-scale text collections inside a parallel data analytics system. Within this topic, we conduct four consecutive research tasks: First, we introduce a set of domain-independent, algebraic operators, which address all fundamental tasks in IE and web text analytics (WA) and which can be used to express complex IE tasks in form of queries inside the declarative data flow language Meteor. Such queries are parsed into algebraic data flows, which are logically and physically optimized, translated into parallel data flow programs, and executed with the parallel processing system Strato- sphere. We characterize all operators with physical, algebraic, and runtime properties to highlight both the potential and importance of optimizing the execution order non- relational, user-defined data flow operators (UDFs). Second, we survey the state-of-the- art in optimization techniques for data flows with UDFs, which are applied at different stages of the optimization process in parallel data analytics systems. We provide a com- prehensive overview on declarative data flow languages for parallel data analytics sys- tems from the perspective of their build-in optimization techniques and conclude that comprehensive optimization of UDFs and non-relational operators still is a true chal- lenge for many systems. Third, we introduce a semantics-aware and extensible logical optimizer for data flows with UDFs based on this observation. Our optimizer builds on a concise set of properties for describing the UDF’s semantics and combines automated analysis of UDFs with manual annotations to enable comprehensive data flow optimiza- tion. We show that our approach is capable of reordering data flows of arbitrary shape from different application domains, leading to considerable runtime improvements and clearly outperforming plans found by other techniques. Fourth, we study the real-life applicability of our system to Terabyte-scale text collections in a challenging setting to compare the "web view" on health-related topics with that derived from a controlled scientific corpus. We systematically evaluate scalability, quality, and robustness of the employed methods and tools and also pinpoint the most critical challenges in building such a system. iv Acknowledgments After a period of almost six years, today is the day: writing this note of thanks is the finishing touch on my dissertation. It has been a period of intense learning for me, not only scientifically, but also on a personal level and I would like to thank the people who have supported me most during this thime. First and foremost, I would like to express my sincere gratitude to my advisor Ulf Leser for his continuous support and encouragement during my academic life over the past 10 years. Ever since I was an undergraduate student, he gave me the opportunity to work on amazing projects, which raised my interest in computer science research in general and in large-scale data management in particular. His guidance, patience, and immense knowledge helped me in all this time and especially during researching and writing this thesis. Thank you, Ulf! I greatly appreaciate the inspiring environment provided by all members of the DFG research group Stratosphere – Information Management on the Cloud. Many thanks to all principal investigators and my fellow graduate students, who listened and engaged in many fruitful discussions during our project meetings. Especially Arvid Heise, Fabian Hueske, and Felix Naumann helped a lot to sharpen and improve results presented here by asking the right questions and by challenging me constantly. I am grateful for the dedication of my student assistants Anja Kunkel, Martin Beckmann, and Jörg Meier, who worked tirelessly to ensure that our system kept up to speed with the development of Stratosphere’s core components. I would also like to thank my fellow graduate students and co-workers at WBI, who made sure that I always enjoyed driving to Adlershof in the morning. I thank Stefan Kröger for being the best office mate I could imagine having for more than five years. Especially spending many coffee breaks with Karin Zimmermann, Philippe Thomas, Marc Bux, and Birgit Heene and chatting about life-related topics provided just the right amount of distraction whenever I needed it. Most importantly, I would like to thank my family for their continuous support and love. Especially Christian, who encouraged
Recommended publications
  • Java Linksammlung
    JAVA LINKSAMMLUNG LerneProgrammieren.de - 2020 Java einfach lernen (klicke hier) JAVA LINKSAMMLUNG INHALTSVERZEICHNIS Build ........................................................................................................................................................... 4 Caching ....................................................................................................................................................... 4 CLI ............................................................................................................................................................... 4 Cluster-Verwaltung .................................................................................................................................... 5 Code-Analyse ............................................................................................................................................. 5 Code-Generators ........................................................................................................................................ 5 Compiler ..................................................................................................................................................... 6 Konfiguration ............................................................................................................................................. 6 CSV ............................................................................................................................................................. 6 Daten-Strukturen
    [Show full text]
  • Declarative Languages for Big Streaming Data a Database Perspective
    Tutorial Declarative Languages for Big Streaming Data A database Perspective Riccardo Tommasini Sherif Sakr University of Tartu Unversity of Tartu [email protected] [email protected] Emanuele Della Valle Hojjat Jafarpour Politecnico di Milano Confluent Inc. [email protected] [email protected] ABSTRACT sources and are pushed asynchronously to servers which are The Big Data movement proposes data streaming systems to responsible for processing them [13]. tame velocity and to enable reactive decision making. However, To facilitate the adoption, initially, most of the big stream approaching such systems is still too complex due to the paradigm processing systems provided their users with a set of API for shift they require, i.e., moving from scalable batch processing to implementing their applications. However, recently, the need for continuous data analysis and pattern detection. declarative stream processing languages has emerged to simplify Recently, declarative Languages are playing a crucial role in common coding tasks; making code more readable and main- fostering the adoption of Stream Processing solutions. In partic- tainable, and fostering the development of more complex appli- ular, several key players introduce SQL extensions for stream cations. Thus, Big Data frameworks (e.g., Flink [9], Spark [3], 1 processing. These new languages are currently playing a cen- Kafka Streams , and Storm [19]) are starting to develop their 2 3 4 tral role in fostering the stream processing paradigm shift. In own SQL-like approaches (e.g., Flink SQL , Beam SQL , KSQL ) this tutorial, we give an overview of the various languages for to declaratively tame data velocity. declarative querying interfaces big streaming data.
    [Show full text]
  • Apache Apex: Next Gen Big Data Analytics
    Apache Apex: Next Gen Big Data Analytics Thomas Weise <[email protected]> @thweise PMC Chair Apache Apex, Architect DataTorrent Apache Big Data Europe, Sevilla, Nov 14th 2016 Stream Data Processing Data Delivery Transform / Analytics Real-time visualization, … Declarative SQL API Data Beam Beam SAMOA Operator SAMOA DAG API Sources Library Events Logs Oper1 Oper2 Oper3 Sensor Data Social Databases CDC (roadmap) 2 Industries & Use Cases Financial Services Ad-Tech Telecom Manufacturing Energy IoT Real-time Call detail record customer facing (CDR) & Supply chain Fraud and risk Smart meter Data ingestion dashboards on extended data planning & monitoring analytics and processing key performance record (XDR) optimization indicators analysis Understanding Reduce outages Credit risk Click fraud customer Preventive & improve Predictive assessment detection behavior AND maintenance resource analytics context utilization Packaging and Improve turn around Asset & Billing selling Product quality & time of trade workforce Data governance optimization anonymous defect tracking settlement processes management customer data HORIZONTAL • Large scale ingest and distribution • Enforcing data quality and data governance requirements • Real-time ELTA (Extract Load Transform Analyze) • Real-time data enrichment with reference data • Dimensional computation & aggregation • Real-time machine learning model scoring 3 Apache Apex • In-memory, distributed stream processing • Application logic broken into components (operators) that execute distributed in a cluster •
    [Show full text]
  • Informatica 10.2 Hotfix 2 Release Notes April 2019
    Informatica 10.2 HotFix 2 Release Notes April 2019 © Copyright Informatica LLC 1998, 2020 Contents Installation and Upgrade......................................................................... 3 Informatica Upgrade Paths......................................................... 3 Upgrading from 9.6.1............................................................. 4 Upgrading from Version 10.0, 10.1, 10.1.1, and 10.1.1 HotFix 1.............................. 4 Upgrading from Version 10.1.1 HF2.................................................. 5 Upgrading from 10.2.............................................................. 6 Related Links ................................................................... 7 Verify the Hadoop Distribution Support................................................ 7 Hotfix Installation and Rollback..................................................... 8 10.2 HotFix 2 Fixed Limitations and Closed Enhancements........................................ 17 Analyst Tool Fixed Limitations and Closed Enhancements (10.2 HotFix 2).................... 17 Application Service Fixed Limitations and Closed Enhancements (10.2 HotFix 2)............... 17 Command Line Programs Fixed Limitations and Closed Enhancements (10.2 HotFix 2).......... 17 Developer Tool Fixed Limitations and Closed Enhancements (10.2 HotFix 2).................. 18 Informatica Connector Toolkit Fixed Limitations and Closed Enhancements (10.2 HotFix 2) ...... 18 Mappings and Workflows Fixed Limitations (10.2 HotFix 2)............................... 18 Metadata
    [Show full text]
  • HDP 3.1.4 Release Notes Date of Publish: 2019-08-26
    Release Notes 3 HDP 3.1.4 Release Notes Date of Publish: 2019-08-26 https://docs.hortonworks.com Release Notes | Contents | ii Contents HDP 3.1.4 Release Notes..........................................................................................4 Component Versions.................................................................................................4 Descriptions of New Features..................................................................................5 Deprecation Notices.................................................................................................. 6 Terminology.......................................................................................................................................................... 6 Removed Components and Product Capabilities.................................................................................................6 Testing Unsupported Features................................................................................ 6 Descriptions of the Latest Technical Preview Features.......................................................................................7 Upgrading to HDP 3.1.4...........................................................................................7 Behavioral Changes.................................................................................................. 7 Apache Patch Information.....................................................................................11 Accumulo...........................................................................................................................................................
    [Show full text]
  • Apache Calcite: a Foundational Framework for Optimized Query Processing Over Heterogeneous Data Sources
    Apache Calcite: A Foundational Framework for Optimized Query Processing Over Heterogeneous Data Sources Edmon Begoli Jesús Camacho-Rodríguez Julian Hyde Oak Ridge National Laboratory Hortonworks Inc. Hortonworks Inc. (ORNL) Santa Clara, California, USA Santa Clara, California, USA Oak Ridge, Tennessee, USA [email protected] [email protected] [email protected] Michael J. Mior Daniel Lemire David R. Cheriton School of University of Quebec (TELUQ) Computer Science Montreal, Quebec, Canada University of Waterloo [email protected] Waterloo, Ontario, Canada [email protected] ABSTRACT argued that specialized engines can offer more cost-effective per- Apache Calcite is a foundational software framework that provides formance and that they would bring the end of the “one size fits query processing, optimization, and query language support to all” paradigm. Their vision seems today more relevant than ever. many popular open-source data processing systems such as Apache Indeed, many specialized open-source data systems have since be- Hive, Apache Storm, Apache Flink, Druid, and MapD. Calcite’s ar- come popular such as Storm [50] and Flink [16] (stream processing), chitecture consists of a modular and extensible query optimizer Elasticsearch [15] (text search), Apache Spark [47], Druid [14], etc. with hundreds of built-in optimization rules, a query processor As organizations have invested in data processing systems tai- capable of processing a variety of query languages, an adapter ar- lored towards their specific needs, two overarching problems have chitecture designed for extensibility, and support for heterogeneous arisen: data models and stores (relational, semi-structured, streaming, and • The developers of such specialized systems have encoun- geospatial). This flexible, embeddable, and extensible architecture tered related problems, such as query optimization [4, 25] is what makes Calcite an attractive choice for adoption in big- or the need to support query languages such as SQL and data frameworks.
    [Show full text]
  • Hortonworks Data Platform for Enterprise Data Lakes Delivers Robust, Big Data Analytics That Accelerate Decision Making and Innovation
    IBM Europe Software Announcement ZP18-0220, dated March 20, 2018 Hortonworks Data Platform for Enterprise Data Lakes delivers robust, big data analytics that accelerate decision making and innovation Table of contents 1 Overview 5 Technical information 2 Key prerequisites 6 Ordering information 2 Planned availability date 7 Terms and conditions 2 Description 9 Prices 5 Program number 10 Announcement countries 5 Publications 10 Corrections Overview Hortonworks Data Platform is an enterprise ready open source Apache Hadoop distribution based on a centralized architecture supported by YARN. Hortonworks Data Platform is designed to address the needs of data at rest, power real-time customer applications, and deliver big data analytics that can help accelerate decision making and innovation. The official Apache versions for Hortonworks Data Platform V2.6.4 include: • Apache Accumulo 1.7.0 • Apache Atlas 0.8.0 • Apache Calcite 1.2.0 • Apache DataFu 1.3.0 • Apache Falcon 0.10.0 • Apache Flume 1.5.2 • Apache Hadoop 2.7.3 • Apache HBase 1.1.2 • Apache Hive 1.2.1 • Apache Hive 2.1.0 • Apache Kafka 0.10.1 • Apache Knox 0.12.0 • Apache Mahout 0.9.0 • Apache Oozie 4.2.0 • Apache Phoenix 4.7.0 • Apache Pig 0.16.0 • Apache Ranger 0.7.0 • Apache Slider 0.92.0 • Apache Spark 1.6.3 • Apache Spark 2.2.0 • Apache Sqoop 1.4.6 • Apache Storm 1.1.0 • Apache TEZ 0.7.0 • Apache Zeppelin 0.7.3 IBM Europe Software Announcement ZP18-0220 IBM is a registered trademark of International Business Machines Corporation 1 • Apache ZooKeeper 3.4.6 IBM(R) clients can download this new offering from Passport Advantage(R).
    [Show full text]
  • Hortonworks Data Platform Date of Publish: 2018-09-21
    Release Notes 3 Hortonworks Data Platform Date of Publish: 2018-09-21 http://docs.hortonworks.com Contents HDP 3.0.1 Release Notes..........................................................................................3 Component Versions.............................................................................................................................................3 New Features........................................................................................................................................................ 3 Deprecation Notices..............................................................................................................................................4 Terminology.............................................................................................................................................. 4 Removed Components and Product Capabilities.....................................................................................4 Unsupported Features........................................................................................................................................... 4 Technical Preview Features......................................................................................................................4 Upgrading to HDP 3.0.1...................................................................................................................................... 5 Before you begin.....................................................................................................................................
    [Show full text]
  • Classifying, Evaluating and Advancing Big Data Benchmarks
    Classifying, Evaluating and Advancing Big Data Benchmarks Dissertation zur Erlangung des Doktorgrades der Naturwissenschaften vorgelegt beim Fachbereich 12 Informatik der Johann Wolfgang Goethe-Universität in Frankfurt am Main von Todor Ivanov aus Stara Zagora Frankfurt am Main 2019 (D 30) vom Fachbereich 12 Informatik der Johann Wolfgang Goethe-Universität als Dissertation angenommen. Dekan: Prof. Dr. Andreas Bernig Gutachter: Prof. Dott. -Ing. Roberto V. Zicari Prof. Dr. Carsten Binnig Datum der Disputation: 23.07.2019 Abstract The main contribution of the thesis is in helping to understand which software system parameters mostly affect the performance of Big Data Platforms under realistic workloads. In detail, the main research contributions of the thesis are: 1. Definition of the new concept of heterogeneity for Big Data Architectures (Chapter 2); 2. Investigation of the performance of Big Data systems (e.g. Hadoop) in virtual- ized environments (Section 3.1); 3. Investigation of the performance of NoSQL databases versus Hadoop distribu- tions (Section 3.2); 4. Execution and evaluation of the TPCx-HS benchmark (Section 3.3); 5. Evaluation and comparison of Hive and Spark SQL engines using benchmark queries (Section 3.4); 6. Evaluation of the impact of compression techniques on SQL-on-Hadoop engine performance (Section 3.5); 7. Extensions of the standardized Big Data benchmark BigBench (TPCx-BB) (Section 4.1 and 4.3); 8. Definition of a new benchmark, called ABench (Big Data Architecture Stack Benchmark), that takes into account the heterogeneity of Big Data architectures (Section 4.5). The thesis is an attempt to re-define system benchmarking taking into account the new requirements posed by the Big Data applications.
    [Show full text]
  • Avro Schema Builder Date
    Avro Schema Builder Date Grove jags moreover. Archie rejoin doubly? Freckly Erasmus magging globularly and stiltedly, she roll-up her berets dieting heritably. Reading the network and avro schema date Processed may be printed on the table into the copier. Kafka avro date and the builder to your experience and avro schema builder date and writers an additional component provides optimizations to. Json Schema Designer Online Clare Locke. It is not read or approved by Pivotal and does not necessarily reflect the views and opinions of Pivotal nor does it constitute any official communication of Pivotal. Post is it in avro schema json to join the kafka takes longer in the data itself and serialization of primitive or information about avro. Meet the apache avro schema defined as the schema requirements change and site design will recall an implementation detail. Data partitioning is critical to data processing performance especially for large volume of data processing in Spark. The Workflow is configured to run daily when new. Date week-millis time-micros timestamp-millis timestamp-micros. Apache avro schema could reject the builder from the files in damages be good http is unknown number of writing. Tracing system collecting latency data from applications. SparkSession val spark SparkSessionbuildermasterlocal. This beard is a beginner's guide my writing came first Avro schema and so few tips for. Entity from avro schemas, dates can or either of the builder for instance of this article, as a schema from the jupyter notebook demonstrates how. Gets builder from schema from a date? Avro date as avro types are you are relevant and analytics query may be very easy for bytes in.
    [Show full text]
  • Informatica® Informatica 10.2 Hotfix 1
    Informatica® Informatica 10.2 HotFix 1 Notas de la versión Informatica Informatica Notas de la versión 10.2 HotFix 1 Agosto 2018 © Copyright Informatica LLC 1998, 2018 Fecha de publicación: 2018-09-25 Tabla de contenido Resumen....................................................................... vi Capítulo 1: Instalación y actualización........................................ 7 Rutas de actualización de Informatica......................................... 7 Cambios en la compatibilidad.............................................. 8 Cambios en la compatibilidad - Distribuciones de Hadoop para Big Data Management....... 9 Cambios en la compatibilidad - Distribuciones de Intelligent Streaming Hadoop.......... 10 Migración a una base de datos diferente....................................... 10 Actualización a la nueva configuración........................................ 10 Actualización desde la versión 10.1.1 HotFix 2................................... 11 Actualizar desde la versión 9.6.1............................................ 11 Vulnerabilidades solucionadas de bibliotecas de otros fabricantes...................... 12 Instalación y reversión de la revisión......................................... 21 Tareas previas a la instalación.......................................... 21 Aplicación o reversión del HotFix en modo gráfico............................. 22 Aplicación o reversión del HotFix en modo de consola........................... 23 Aplicación o reversión del HotFix en modo silencioso........................... 24 Aplicación
    [Show full text]
  • Code Smell Prediction Employing Machine Learning Meets Emerging Java Language Constructs"
    Appendix to the paper "Code smell prediction employing machine learning meets emerging Java language constructs" Hanna Grodzicka, Michał Kawa, Zofia Łakomiak, Arkadiusz Ziobrowski, Lech Madeyski (B) The Appendix includes two tables containing the dataset used in the paper "Code smell prediction employing machine learning meets emerging Java lan- guage constructs". The first table contains information about 792 projects selected for R package reproducer [Madeyski and Kitchenham(2019)]. Projects were the base dataset for cre- ating the dataset used in the study (Table I). The second table contains information about 281 projects filtered by Java version from build tool Maven (Table II) which were directly used in the paper. TABLE I: Base projects used to create the new dataset # Orgasation Project name GitHub link Commit hash Build tool Java version 1 adobe aem-core-wcm- www.github.com/adobe/ 1d1f1d70844c9e07cd694f028e87f85d926aba94 other or lack of unknown components aem-core-wcm-components 2 adobe S3Mock www.github.com/adobe/ 5aa299c2b6d0f0fd00f8d03fda560502270afb82 MAVEN 8 S3Mock 3 alexa alexa-skills- www.github.com/alexa/ bf1e9ccc50d1f3f8408f887f70197ee288fd4bd9 MAVEN 8 kit-sdk-for- alexa-skills-kit-sdk- java for-java 4 alibaba ARouter www.github.com/alibaba/ 93b328569bbdbf75e4aa87f0ecf48c69600591b2 GRADLE unknown ARouter 5 alibaba atlas www.github.com/alibaba/ e8c7b3f1ff14b2a1df64321c6992b796cae7d732 GRADLE unknown atlas 6 alibaba canal www.github.com/alibaba/ 08167c95c767fd3c9879584c0230820a8476a7a7 MAVEN 7 canal 7 alibaba cobar www.github.com/alibaba/
    [Show full text]