Pre-classification and AI

Slides:



Advertisements
Similar presentations
Chapter 5: Introduction to Information Retrieval
Advertisements

Decision Tree Approach in Data Mining
New Technologies Supporting Technical Intelligence Anthony Trippe, 221 st ACS National Meeting.
© Devon M.Simmonds, 2007 CSC 550 Graduate Course in Software Engineering ______________________ Devon M. Simmonds Computer Science Department University.
Benjamin J. Deaver Advisor – Dr. LiGuo Huang Department of Computer Science and Engineering Southern Methodist University.
SBSE Course 3. EA applications to SE Analysis Design Implementation Testing Reference: Evolutionary Computing in Search-Based Software Engineering Leo.
The Decision-Making Process IT Brainpower
11.1 Lecture 11 CASE tools IMS Systems Design and Implementation.
Information Retrieval Concerned with the: Representation of Storage of Organization of, and Access to Information items.
Evolving Neural Networks in Classification Sunghwan Sohn.
2000 International Conference on Engineering Education1 The Web-Based Learning Environment for Creative Design Course S. S. Hsiau, J. C. Wu, T. L. Yeh.
Review 4 Chapters 8, 9, 10.
Configuration Management IACT 918 July 2004 Gene Awyzio SITACS University of Wollongong.
Report on Multi-agent Data Fusion System: Design and implementation issues 1 By Ganesh Godavari.
Configuration Management IACT 418/918 Autumn 2005 Gene Awyzio SITACS University of Wollongong.
Introduction to Systems Analysis and Design
©Ian Sommerville 2004Software Engineering, 7th edition. Chapter 8 Slide 1 Tools of Software Development l 2 types of tools used by software engineers:
Software Testing Test Design and Implementation. Agenda Test Design Test Implementation Test Design Sources Automated Testing 2.
1 Text Categorization  Assigning documents to a fixed set of categories  Applications:  Web pages  Recommending pages  Yahoo-like classification hierarchies.
0 © WIPO – 2003 PF & CJF CLAIMS Computer-Assisted Categorisation of Patent Documents in the International Patent Classification Patrick Fiévet, CLAIMS.
Language Identification of Search Engine Queries Hakan Ceylan Yookyung Kim Department of Computer Science Yahoo! Inc. University of North Texas 2821 Mission.
Introduction to SDLC: System Development Life Cycle Dr. Dania Bilal IS 582 Spring 2009.
1SAS 03/ GSFC/SATC- NSWC-DD System and Software Reliability Dolores R. Wallace SRS Technologies Software Assurance Technology Center
Automated Patent Classification By Yu Hu. Class 706 Subclass 12.
The Asset Inventory Management module assists with data collection and discovery management processes. Collected information is interpreted and automatically.
11.1 © 2007 by Prentice Hall 11 Chapter Building Information Systems.
11 C H A P T E R Artificial Intelligence and Expert Systems.
Chapter 1 Introduction to Data Mining
Computerised Air Traffic Management Tools - Benefits and Limitations OMAR BASHIR (March 2005)
2Object-Oriented Analysis and Design with the Unified Process The Requirements Discipline in More Detail  Focus shifts from defining to realizing objectives.
Implementation of the reformed IPC: Slovenia Janez Kukec Mezek Head of Information and Promotion Department Slovenian Intellectual Property Office.
Project 1: Machine Learning Using Neural Networks Ver 1.1.
Ensembles. Ensemble Methods l Construct a set of classifiers from training data l Predict class label of previously unseen records by aggregating predictions.
Hendrik J Groenewald Centre for Text Technology (CTexT™) Research Unit: Languages and Literature in the South African Context North-West University, Potchefstroom.
Text Document Categorization by Term Association Maria-luiza Antonie Osmar R. Zaiane University of Alberta, Canada 2002 IEEE International Conference on.
June 13-15, 2007Policy 2007 Infrastructure-aware Autonomic Manager for Change Management H. Abdel SalamK. Maly R. MukkamalaM. Zubair Department of Computer.
Your Interactive Guide to the Digital World Discovering Computers 2012 Chapter 12 Exploring Information System Development.
Automatic Categorization of Patent Applications Presentation to the 3rd IPC Workshop, WIPO, Feb , The need for automatic categorization of.
1 Text Categorization  Assigning documents to a fixed set of categories  Applications:  Web pages  Recommending pages  Yahoo-like classification hierarchies.
A Decision Support Based on Data Mining in e-Banking Irina Ionita Liviu Ionita Department of Informatics University Petroleum-Gas of Ploiesti.
C. Mugnier, D. Lafarge, C. Perolini, R. Pilon, J. Ruiz-Cabezas
IPC reform 2006: WIPO Products and Services for the new IPC Special seminar for patent information vendors World Intellectual Property Organization WIPO.
Introduction to Systems Analysis and Design
Software Defects Cmpe 550 Fall 2005
System.
Applying Deep Neural Network to Enhance EMPI Searching
Application of Classification and Clustering Methods on mVoC (Medical Voice of Customer) data for Scientific Engagement Yingzi Xu, Department of Statistics,
Building Information Systems
Head, IT Systems Section
Tools of Software Development
Head, IT Systems Section
ROADMAP TO ISO/TS REGISTRATION
Project 1: Text Classification by Neural Networks
9.a Report on IPC-related IT systems IPC Committee of Experts 50
Text Categorization Assigning documents to a fixed set of categories
Artificial Intelligence applied to IPC and Nice classifications
CSE 635 Multimedia Information Retrieval
Web Mining Department of Computer Science and Engg.
RNG Implementation Release 1.
Report on IPC-related IT systems IPC Revision Working Group 39
2016 AES – Draft Commission Regulation implementing Regulation (EC) No 452/2008 Agenda item 2.3 DSS Meeting 3-4 April 2014.
ENCODING TOOL DEVELOPED BY HUNGARY Márta Záhonyi
CLAIMS CLassification Automated InforMation System
Deep SEARCH 9 A new tool in the box for automatic content classification: DS9 Machine Learning uses Hybrid Semantic AI ConTech November.
©Ian Sommerville 2004Software Engineering, 7th edition. Chapter 8 Slide 1 Tools of Software Development l 2 types of tools used by software engineers:
FREERIDE: A Framework for Rapid Implementation of Datamining Engines
Project Title This is a sample poster layout -
Active AI Projects at WIPO
Presentation transcript:

Pre-classification and AI French Patent Department

Current procedure Currently: Manual sorting by managers Expectation: Reducing sorting time and limiting distribution errors DB Sorting 1 S1 S2 S3 Sorting 2 P1 P2 P3 - Examiner 1 - Examiner 2 - Examiner 3 - … Automatisation of sorting on levels 1 et 2 by using only classification (subclass level) Managers keep the last sorting task (examiner level) : based on IPC, but also on workload, agendas … 21/01/2019 / Pre-classification and AI

2 possibilities Patent application Computer-Assisted Classification Pre-classification tool 3 or more potential classes (with level of confidence) Most likely examiners’ team Decision support for managers, who will continue to distribute Automatic distribution to this examiners’ team 21/01/2019 / Pre-classification and AI

Methodology Learning : INPI databases (2008-2018) 140 000 applications Patent application - Main IPC Data collection IPC Team Patent application (xml) with associated team Data fusion 21/01/2019 / Pre-classification and AI

Methodology Text processing Model construction Text Mining : lowercase, punctuation removal, stopwords removal, lemmatisation, stemming,… Model construction Machine learning model: FastText In the future, maybe TF-IDF and neural networks Model application on the training set (ca. 20 000 applications) Comparison with the correct team Graphic interface R-Shiny to apply the model on new applications Export of the predicted team 21/01/2019 / Pre-classification and AI

Results Reliability: 85 % of the training set P1 P2 P3 P4 P5 P6 P7 P8 Error rate with manual distribution : 5 to 10 % Error distribution with Computer-Assisted pre-classification Reference P1 P2 P3 P4 P5 P6 P7 P8 1385 51 61 20 49 35 6 36 1503 84 15 39 140 26 56 2160 50 12 21 38 2219 106 24 47 42 66 136 1326 34 14 104 29 25 1589 45 7 9 23 10 1908 76 33 46 43 54 59 1398 Prediction 21/01/2019 / Pre-classification and AI

Recent developpments (6 months) Focusing points Impact analysis Testing phase Recent developpments (6 months) Focusing points Impact analysis Processing delays in case of error Reattribution mechanism Stabilisation and implementation of the model IPC Proposal for examiners 21/01/2019 / Pre-classification and AI