machine learning vs knowledge based approaches to adr
play

Machine learning vs. knowledge based approaches to ADR - PowerPoint PPT Presentation

May 10, 2023 •22 likes •242 views

Machine learning vs. knowledge based approaches to ADR identification November 2017 Topics Short about us Iden3fying ADRs Machine Learning for seman3c rela3ons iden3fica3on Results Challenges SHORT ABOUT US Focus on

  1. Machine learning vs. knowledge based approaches to ADR identification November 2017

  2. Topics Short about us • Iden3fying ADRs • Machine Learning for seman3c rela3ons • iden3fica3on Results • Challenges •

  3. SHORT ABOUT US Focus on text-analy3cs for Pharmaceu3cals. Since 1998 Othe her text xt sou sources s Voi oice of of the he Pa3 Pa3ent nt Electroni Ele onic He Healt alth h Re Recor ords ds Scien3fic literature – FDA – Patents – Business opportuni3es

  4. SHORT ABOUT US Advanced Databases Group, Universidad Carlos III de Madrid § Research lines: • Natural language processing • Accessibility § Resources produced: • Drug-drug-interac-on collec-on (DDI Corpus) • DINTO ontology

  5. Our goal at TAC ADR “ Combine Knowledge Based with Machine Learning Based approaches to leverage ADR identification ”

  6. Iden%fying ADRs

  7. TO TOPI PIC C EXTRA EXTRACTI CTION N NLP and Resource based approach ¡ SI SIDER DER ¡ UMLS UMLS ¡ Tr Trai aini ning ng cor orpus s

  8. TO TOPI PIC C EXTRA EXTRACTI CTION N NLP and Resource based approach ¡ SI SIDER DER ¡ UMLS UMLS ¡ Tr Trai aini ning ng cor orpus s Dictionary #entries Adverse Reactions 21,826 Factor 41 Severity 158 Animal 27 DrugClass 101

  9. TO TOPI PIC C EXTRA EXTRACTI CTION N NLP and Resource based approach And some rules to iden3fy nega3on: • MeaningCloud Insights Engine API supports this rule syntax •

  10. Machine Learning for seman%c rela%ons iden%fica%on

  11. Machine learning for seman3c rela3ons iden3fica3on Represen3ng ADR men3on context through a set of features: ADRMen9on – Other pairs (where Other is Severity, DrugClass, Ø M1TXT, M1TXT, M2TXT, M2TXT, BWTXT BWTXT: the text of both/between men3ons. Nega9on, Animal or Factor) Ø C1BO C1BOW, W, C2BO C2BOW: bag-of-words of both men3ons. Ø C1PO C1POS, S, C2PO C2POS: part of speech of both men3ons. Ø PB1PO PB1POS, S, PA PA1PO 1POS, S, PB2PO PB2POS, S, PA PA2PO 2POS, S, PWPO PWPOS: the PoS tags of the two tokens before/ aZer/between both men3ons. Ø WA WA1TXT, 1TXT, WB2TXT, WB2TXT, WA WA2TXT, 2TXT, WB1TXT WB1TXT: the two tokens aZer/before the men3on. Ø LA LA1LEM, 1LEM, LB2LEM, LB2LEM, LA LA2LEM, 2LEM, LB1LEM LB1LEM: the lemmas of the two tokens aZer/before both men3ons. Ø LWLEM LWLEM: the lemmas between of the two men3ons Ø NTO NTOKB KB: the number of tokens between the two men3ons.

  12. Machine learning for seman3c rela3ons iden3fica3on And the algorithm? § SVM, support vector machines (using scikit-learn on Python) § Specifically, SVC implementa3on: • Default parameter values • Linear kernel But, no deep learning??!! Of course (CNN), but not in the official runs.

  13. Results

  14. Results • Task 1. ADR and related en33es Low precision!! • Task 2. Rela3ons between ADRs and en33es Oh, oh!!

  15. Results • Task 3. Posi3ve ADRs Only a few negated men9ons? PreQy good!! • Task 4. Normaliza3on through MedDRA Using dic9onaries with seman9c informa9on produces nice results

  16. Challenges

  17. Challenges • Nega3on iden3fica3on requires more effort (not only on the ADRs field). Some weird things found in the test set: Eg.: The most frequently observed malignancies other than non- melanoma skin cancer … Negation? • CNNs and the use of syntac3c features improves results P R F1 Other 0.71 0.81 0.76 Negated 0.72 0.40 0.51 Hypothetical 0.75 0.75 0.75 Effect 0.76 0.61 0.68 Avg / total 0.73 0.73 0.73

  18. Challenges • Recall must be improved: o separated mul3word men3ons o ADRs with no MedDRA code, enough lexical resources? • How to approach errors when applying deep learning? • Enough accuracy for prac3cal applica3ons? What does FDA think?

  19. Thanks QUESTIONS?

  20. LabDA Resources Corpus DDI (Drug-Drug Interac3ons) 1,025 annotated documents, 18,502 en33es and } 5,028 DDIs (by expert pharma) MedLine and DrugBank texts } Annota3ns guidelines and interannotator agreement. } Available at labda.inf.uc3m.es } Used at DDIExtrac3on 2011 and DDIExtrac3on 2013 } Semeval Tasks

  21. LabDA Resources DINTO Ontology- knowledge about drugs and intera3ons (11,555 DDIs and 8,786 pharmacological en33es). Available at OBO Foundry Applica3on to Informa3on Extrac3on and Predic3on

  22. Meani Me aning ngClou Cloud d LLC LLC Automa3ng the extrac3on of Meaning from any informa3on source. Addr ddress ss 35-37 36th Street 11106 Astoria NY Cont Contac act Inf nfo o jmar3nez@meaningcloud.com Tele Te lephone hone Phone: +1 (646) 403-3104 meaningcloud.com

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

10a Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10a Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10a Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A Framework for 10.6 Unsupervised Learning Symbol-based Learning 10.7 Reinforcement Learning 10.2 Version Space Search 10.8 Epilogue and 10.3 The

934 views • 54 slides
10d Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10d Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10d Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A Framework for 10.6 Unsupervised Learning Symbol-based Learning 10.7 Reinforcement Learning 10.2 Version Space Search 10.8 Epilogue and 10.3 The

1.05k views • 29 slides
10c Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10c Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10c Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A Framework for 10.6 Unsupervised Learning Symbol-based Learning 10.7 Reinforcement Learning 10.2 Version Space Search 10.8 Epilogue and 10.3 The

605 views • 35 slides
10b Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10b Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A

10b Machine Learning: Symbol-based 10.0 Introduction 10.5 Knowledge and Learning 10.1 A Framework for 10.6 Unsupervised Learning Symbol-based Learning 10.7 Reinforcement Learning 10.2 Version Space Search 10.8 Epilogue and 10.3 The

866 views • 50 slides
Machine-Learning & AI-based Approaches for GPCR Bioactive Ligand Discovery Sebastian Raschka,

Machine-Learning & AI-based Approaches for GPCR Bioactive Ligand Discovery Sebastian Raschka,

Cambridge Healthtech Institute's 14th Annual GPCR-Based Drug Discovery Discovery on Target Conference September 19, 2019 | Boston, MA https://www.discoveryontarget.com/GPCR-drug-discovery Machine-Learning & AI-based Approaches for GPCR

411 views • 38 slides
TAXONOMY AND CHALLENGES IN MACHINE LEARNING-BASED APPROACHES TO DETECT ATTACKS IN THE INTERNET

TAXONOMY AND CHALLENGES IN MACHINE LEARNING-BASED APPROACHES TO DETECT ATTACKS IN THE INTERNET

The 15th International Conference on Availability, Reliability and Security (ARES 2020) August 25 to August 28, 2020 in Dublin, Ireland ID-86 workshop paper (IoT-SECFOR) TAXONOMY AND CHALLENGES IN MACHINE LEARNING-BASED APPROACHES TO DETECT

583 views • 13 slides
Linguistics 384: Language and Computers approaches Linguistic knowledge Topic 5: Machine

Linguistics 384: Language and Computers approaches Linguistic knowledge Topic 5: Machine

Language and Computers Topic 5: Machine Translation Introduction Examples for Translations Background: Dictionaries Transformer Linguistics 384: Language and Computers approaches Linguistic knowledge Topic 5: Machine Translation based

1.85k views • 184 slides
Machine Learning: Der KDD-Prozess Knowledge Discovery in Databases Machine Learning Data-Mining

Machine Learning: Der KDD-Prozess Knowledge Discovery in Databases Machine Learning Data-Mining

Machine Learning: Der KDD-Prozess Knowledge Discovery in Databases Machine Learning Data-Mining Aufgaben Aufgabe Von Daten zum Wissen Wissensextraktion durch We are drowing in automatisches Erkennen von information,

1.01k views • 47 slides
Learning Approaches to Estimate Depth from RGB Lecture 5 What will we learn - Latest Approaches

Learning Approaches to Estimate Depth from RGB Lecture 5 What will we learn - Latest Approaches

Learning Approaches to Estimate Depth from RGB Lecture 5 What will we learn - Latest Approaches to Depth Estimation based on Machine Learning (DNNs) Why do we need new approaches? Paper1 -> CNNs for Depth estimation Paper2

634 views • 17 slides
The logic of learning: The logic of learning: logic and knowledge representation logic and

The logic of learning: The logic of learning: logic and knowledge representation logic and

The logic of learning: The logic of learning: logic and knowledge representation logic and knowledge representation in machine learning in machine learning Peter A. Flach Department of Computer Science University of Bristol

771 views • 59 slides
Linguistics 384: Language and Computers Transformer approaches Topic 5: Machine Translation

Linguistics 384: Language and Computers Transformer approaches Topic 5: Machine Translation

Language and Computers Topic 5: Machine Translation Introduction Examples for Translations Background: Dictionaries Linguistics 384: Language and Computers Transformer approaches Topic 5: Machine Translation Linguistic knowledge based

2.2k views • 184 slides
Knowledge Engineering Sargur Srihari srihari@cedar.buffalo.edu 1 Machine Learning Srihari

Knowledge Engineering Sargur Srihari srihari@cedar.buffalo.edu 1 Machine Learning Srihari

Machine Learning Srihari Knowledge Engineering Sargur Srihari srihari@cedar.buffalo.edu 1 Machine Learning Srihari Topics Picking Variables Determining Structure Determining Probabilities 2 Machine Learning Srihari Knowledge

313 views • 12 slides
26:198:722 Expert Systems I Knowledge representation I Knowledge acquisition I Machine learning I

26:198:722 Expert Systems I Knowledge representation I Knowledge acquisition I Machine learning I

26:198:722 Expert Systems I Knowledge representation I Knowledge acquisition I Machine learning I ID3 & C4.5 Knowledge Representation Recall: I Knowledge engineering F Knowledge acquisition N Knowledge elicitation F Knowledge representation N

668 views • 62 slides
I. Evolutionary Knowledge Machine (EKM) II. AI Doctor Platform (IMP) III. Case Studies (Silo)

I. Evolutionary Knowledge Machine (EKM) II. AI Doctor Platform (IMP) III. Case Studies (Silo)

AI Doctor based on Department of Computer Science & Engineering, Evolutionary Knowledge Machine (EKM) Kyung Hee University KOREA February 19 th , 2019 Prof. Sungyoung Lee http://uclab.khu.ac.kr 2 Contents / I. Evolutionary Knowledge

416 views • 41 slides
Genetics-based Machine Learning and Behaviour Based Robotics: A New Synthesis

Genetics-based Machine Learning and Behaviour Based Robotics: A New Synthesis

Genetics-based Machine Learning and Behaviour Based Robotics: A New Synthesis Genetics-based Machine Learning and Behavior Based Robotics: A New Synthesis, Marco Dorigo, Uwe Schnepf, IEEE Transactions on System, MAn, and Cybernetics,

348 views • 34 slides
Interpreting Embedding Models of Knowledge Bases: Model Agnostic Approaches 2018 ICML Workshop on

Interpreting Embedding Models of Knowledge Bases: Model Agnostic Approaches 2018 ICML Workshop on

Introduction & Background Interpreting Embedding Models of KBs Experiments Conclusion Interpreting Embedding Models of Knowledge Bases: Model Agnostic Approaches 2018 ICML Workshop on Human Interpretability in Machine Learning Arthur C.

751 views • 24 slides
effective as a supervised exercise training program for people with CVD? Hazel Mountford &

effective as a supervised exercise training program for people with CVD? Hazel Mountford &

Is home-based exercise training as effective as a supervised exercise training program for people with CVD? Hazel Mountford & Madeline Gaynor Physiotherapy Department Sir Charles Gairdner Hospital, Western Australia Background

817 views • 23 slides
reduces susceptibility to pulmonary arterial hypertension in bmpr2 deficient mice A Crosby, E

reduces susceptibility to pulmonary arterial hypertension in bmpr2 deficient mice A Crosby, E

Bone marrow transplantation reduces susceptibility to pulmonary arterial hypertension in bmpr2 deficient mice A Crosby, E Soon, M Southwood, M Toshner, BJ Dunmore, NW Morrell Pulmonary arterial hypertension PAH (PAP >25 mmHg)

952 views • 72 slides
CD38 and Ceramide-Enriched Membrane Platforms in Bovine Coronary Myocytes July 31, 2008 Matt

CD38 and Ceramide-Enriched Membrane Platforms in Bovine Coronary Myocytes July 31, 2008 Matt

CD38 and Ceramide-Enriched Membrane Platforms in Bovine Coronary Myocytes July 31, 2008 Matt Childrey Department of Pharmacology & Toxicology Medical College of Virginia Virginia Commonwealth University Richmond, Virginia Summer Program

155 views • 13 slides
City of Racine 2020 High Deductible HSA Compatible Health Plan Overview Agenda Benefit

City of Racine 2020 High Deductible HSA Compatible Health Plan Overview Agenda Benefit

City of Racine 2020 High Deductible HSA Compatible Health Plan Overview Agenda Benefit Plan Overview Benefit Plan Considerations HSA Education & Guidelines Medicare Advantage Plan Option 2 2020 PROPOSED Health Plan Design

243 views • 21 slides
Session 3 : Basis for Regulatory Decision-Making Objectives of paediatric developm ent EMA/ FDA/

Session 3 : Basis for Regulatory Decision-Making Objectives of paediatric developm ent EMA/ FDA/

Session 3 : Basis for Regulatory Decision-Making Objectives of paediatric developm ent EMA/ FDA/ Health Canada Workshop on Paediatric Pulmonary Arterial Hypertension London, June 2017 Presented by: Amany El-Gazayerly, Dutch Medicines

489 views • 15 slides
Philip Oddie & Abhishek Oswal The clinical scenario The patient & problem Mrs C.

Philip Oddie & Abhishek Oswal The clinical scenario The patient & problem Mrs C.

Philip Oddie & Abhishek Oswal The clinical scenario The patient & problem Mrs C. Hocolate is a 57 year old known hypertensive who read in the Daily Mail that eating dark chocolate will help reduce her blood pressure The key

314 views • 9 slides
Nutritional Management of Cows Zach McCracken Suther Feeds, Inc. Who We Are Second generation,

Nutritional Management of Cows Zach McCracken Suther Feeds, Inc. Who We Are Second generation,

Opportunity Solution Success Nutritional Management of Cows Zach McCracken Suther Feeds, Inc. Who We Are Second generation, family owned livestock nutrition company started in 1963 Business model - Opportunities : program diversity -

325 views • 13 slides
Lexaria Bioscience Corp. Investor Presentation LXX:CSE | LXRP:US June 2017

Lexaria Bioscience Corp. Investor Presentation LXX:CSE | LXRP:US June 2017

Lexaria Bioscience Corp. Investor Presentation LXX:CSE | LXRP:US June 2017 www.LexariaBioscience.com 1 Disclaimer This presentation includes forward-looking statements within the meaning of Section 27A of the Securities Act of 1933, as

362 views • 15 slides

More recommend