Specialised vs Declarative Data Mining Software Testing - PowerPoint PPT Presentation

EXAMPLE θ = 3 D (2 I , ⊆ ) Closedness M θ = { P ∈ I| freq ( P ) ≥ θ ∧ ∀ P 0 ⊃ P : freq ( P 0 ) < θ } � 10

EXAMPLE θ = 3 D (2 I , ⊆ ) Closedness M θ = { P ∈ I| freq ( P ) ≥ θ ∧ ∀ P 0 ⊃ P : freq ( P 0 ) < θ } � 11

CONDENSED REPRESENTATION � 12

CONDENSED REPRESENTATION Dataset #Frequent #Closed #Maximal 151 807 3 292 230 Zoo-1 Mushroom 155 734 3 287 453 9 967 402 46 802 5 191 Lymph 27 . 10 7 1 827 264 189 205 Hepa;;s � 12

SPECIALIZED VS DECLARATIVE DATA MINING � 13

SPECIALIZED VS DECLARATIVE DATA MINING dataset � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints dataset � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints + Specialised Miner dataset � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints + Specialised Miner Patterns dataset � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints + Specialised Miner Patterns dataset Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints Sophisticated user’s + constraints Specialised Miner Patterns dataset Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints Sophisticated user’s + constraints Specialised Miner 1 Patterns dataset preprocessing Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints Sophisticated user’s 2 + constraints Specialised post- Miner processing 1 Patterns dataset preprocessing Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query new algo 3 Basic user’s constraints Sophisticated user’s 2 + constraints Specialised post- Miner processing 1 Patterns dataset preprocessing Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query new algo 3 Basic user’s constraints Sophisticated user’s 2 + constraints Specialised post- Miner processing 1 Patterns dataset preprocessing Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] Need: Declarative way to deal with more complex queries ➤ Declarative data Mining � 13

SPECIALIZED VS DECLARATIVE DATA MINING Query new algo 3 Basic user’s constraints Sophisticated user’s 2 + CP model constraints post- + CP solver processing 1 Patterns dataset preprocessing Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] Need: Declarative way to deal with more complex queries ➤ Declarative data Mining � 14

SPECIALIZED VS DECLARATIVE DATA MINING Query Basic user’s constraints Sophisticated user’s + CP model constraints + CP solver Patterns dataset Limitations: Dealing with sophisticated user’s constraints [Wojciechowski and Zakrzewicz, 02] Need: Declarative way to deal with more complex queries ➤ Declarative data Mining � 14

SPECIALISED VS DECLARATIVE DATA MINING � 15

SPECIALISED VS DECLARATIVE DATA MINING Specialised is the winner! � 15

SPECIALISED VS DECLARATIVE DATA MINING Specialised is the winner! Declarative is the winner! � 15

SPECIALISED VS DECLARATIVE DATA MINING Preprocessing + Specialised step vs Declarative � 16

SPECIALISED VS DECLARATIVE DATA MINING Specialised + postprocessing vs Declarative � 17

CONCLUSIONS (PART I) � 18

CONCLUSIONS (PART I) ➤ Specialised methods are suitable for: ➤ Enumerating Patterns ➤ Taking into account classic constraints (simple queries) � 18

CONCLUSIONS (PART I) ➤ Specialised methods are suitable for: ➤ Enumerating Patterns ➤ Taking into account classic constraints (simple queries) ➤ Declarative methods are suitable for: ➤ Taking into account user’s constraints (complex queries) ➤ Iterative data mining process � 18

CONCLUSIONS (PART I) ➤ Specialised methods are suitable for: ➤ Enumerating Patterns ➤ Taking into account classic constraints (simple queries) ➤ Declarative methods are suitable for: ➤ Taking into account user’s constraints (complex queries) ➤ Iterative data mining process Time left? � 18

FAULT LOCALISATION � 19

FAULT LOCALISATION ➤ The need: identify a subset of statements that are susceptible to explain a fault in a program ➤ Precision <=> Efficiency � 19

FAULT LOCALISATION ➤ The need: identify a subset of statements that are susceptible to explain a fault in a program ➤ Precision <=> Efficiency ➤ Spectrum-based approaches: (ranking metrics - suspiciousness score) ➤ Tarantula [Jones and Harrold 05] ➤ Ochiai [Abreu et al. 07] ➤ Jaccard [Abreu et al. 07] ➤ … � 19

FAULT LOCALISATION (MOTIVATIONS) � 20

FAULT LOCALISATION (MOTIVATIONS) ➤ Pros: Quick localisation � 20

FAULT LOCALISATION (MOTIVATIONS) ➤ Pros: Quick localisation ➤ Cons: independent evaluation of each statement at the expense of accuracy � 20

FAULT LOCALISATION (MOTIVATIONS) Test cases Program : Character counter tc 1 tc 2 tc 3 tc 4 tc 5 tc 6 tc 7 tc 8 function count (char *s) { int let, dig, other, i = 0; char c; e 1 : while (c = s[i++]) { 1 1 1 1 1 1 1 1 e 2 : 1 1 1 1 1 1 0 1 if(’A’<=c && ’Z’>=c) e 3 : 1 1 1 1 1 1 0 0 let += 2; //- fault - e 4 : 1 1 1 1 1 0 0 1 else if ( ’a’<=c && ’z’>=c ) e 5 : 1 1 0 0 1 0 0 0 let += 1; e 6 : 1 1 1 1 0 0 0 1 else if ( ’0’<=c && ’9’>=c ) e 7 : 0 1 0 1 0 0 0 0 dig += 1; e 8 : 1 0 1 0 0 0 0 1 else if (isprint (c)) e 9 : 1 0 1 0 0 0 0 1 other += 1; e 10 : printf("%d %d %d \ n", let, dig, other); } 1 1 1 1 1 1 1 1 Passing/Failing F F F F F F P P � 21

FAULT LOCALISATION (MOTIVATIONS) ➤ Pros: Quick localisation � 22

FAULT LOCALISATION (MOTIVATIONS) ➤ Pros: Quick localisation ➤ Cons: independent evaluation of each statement at the expense of accuracy � 22

FAULT LOCALISATION (MOTIVATIONS) ➤ Pros: Quick localisation ➤ Cons: independent evaluation of each statement at the expense of accuracy ➤ Need: more finer-grained localisation, taking into account user’s constraints � 22

FAULT LOCALISATION (MOTIVATIONS) ➤ Pros: Quick localisation ➤ Cons: independent evaluation of each statement at the expense of accuracy ➤ Need: more finer-grained localisation, taking into account user’s constraints ➤ How: Use of Declarative Data Mining � 22

FAULT LOCALISATION (MOTIVATIONS) Test cases Program : Character counter tc 1 tc 2 tc 3 tc 4 tc 5 tc 6 tc 7 tc 8 function count (char *s) { int let, dig, other, i = 0; char c; e 1 : while (c = s[i++]) { 1 1 1 1 1 1 1 1 e 2 : 1 1 1 1 1 1 0 1 if(’A’<=c && ’Z’>=c) e 3 : 1 1 1 1 1 1 0 0 let += 2; //- fault - e 4 : 1 1 1 1 1 0 0 1 else if ( ’a’<=c && ’z’>=c ) e 5 : 1 1 0 0 1 0 0 0 let += 1; e 6 : 1 1 1 1 0 0 0 1 else if ( ’0’<=c && ’9’>=c ) e 7 : 0 1 0 1 0 0 0 0 dig += 1; e 8 : 1 0 1 0 0 0 0 1 else if (isprint (c)) e 9 : 1 0 1 0 0 0 0 1 other += 1; e 10 : printf("%d %d %d \ n", let, dig, other); } 1 1 1 1 1 1 1 1 Passing/Failing F F F F F F P P � 23

FAULT LOCALISATION (MOTIVATIONS) Test cases Program : Character counter tc 1 tc 2 tc 3 tc 4 tc 5 tc 6 tc 7 tc 8 function count (char *s) { int let, dig, other, i = 0; char c; e 1 : while (c = s[i++]) { 1 1 1 1 1 1 1 1 e 2 : 1 1 1 1 1 1 0 1 if(’A’<=c && ’Z’>=c) e 3 : 1 1 1 1 1 1 0 0 Fault localisation let += 2; //- fault - e 4 : 1 1 1 1 1 0 0 1 else if ( ’a’<=c && ’z’>=c ) = e 5 : 1 1 0 0 1 0 0 0 let += 1; Mining Task e 6 : 1 1 1 1 0 0 0 1 else if ( ’0’<=c && ’9’>=c ) e 7 : 0 1 0 1 0 0 0 0 dig += 1; e 8 : 1 0 1 0 0 0 0 1 else if (isprint (c)) e 9 : 1 0 1 0 0 0 0 1 other += 1; e 10 : printf("%d %d %d \ n", let, dig, other); } 1 1 1 1 1 1 1 1 Passing/Failing F F F F F F P P � 23

PATTERN SUSPICIOUSNESS DEGREE (PSD) � 24

PATTERN SUSPICIOUSNESS DEGREE (PSD) ➤ PSD function. Given a pattern P of a program: PSD ( P ) = freq − ( P ) + | F AIL | − freq + ( P ) | P ASS | +1 � 24

PATTERN SUSPICIOUSNESS DEGREE (PSD) ➤ PSD function. Given a pattern P of a program: PSD ( P ) = freq − ( P ) + | F AIL | − freq + ( P ) | P ASS | +1 ➤ PSD-dominance relation. Given two patterns P i and P j P i B P SD P j ⇔ PSD ( P i ) > PSD ( P j ) � 24

PATTERN SUSPICIOUSNESS DEGREE (PSD) ➤ PSD function. Given a pattern P of a program: PSD ( P ) = freq − ( P ) + | F AIL | − freq + ( P ) | P ASS | +1 ➤ PSD-dominance relation. Given two patterns P i and P j P i B P SD P j ⇔ PSD ( P i ) > PSD ( P j ) ➤ Top-k suspicious patterns. top-k= { P | 6 9 P 1 , . . . , P k : 8 1  j  k, P j B P SD P } � 24

FCP-MINER TOOL (SOME RESULTS) � 25

CONCLUSIONS (PART II) � 26

Specialised vs Declarative Data Mining Software Testing - PowerPoint PPT Presentation

Specialised vs Declarative Data Mining Software Testing Applications Nadjib Lazaar , CNRS, University of Montpellier Join works with: M. Maamar, Y. Lebbah, S. Loudni, C. Bessiere, et. al. SIMULA, Oslo, 11 oct. 2018 DATA MINING 2 DATA

Analyzing manuscript traditions using constraint-based data mining A case study in declarative

DATA MINING LECTURE 2 What is data? The data mining pipeline What is Data Mining? Data

Web Mining Web Mining Web Mining Web Mining Web mining is the use of data mining techniques

LAB-07 Declarative Process Modeling and Mining Lecturer: Andrea MARRELLA Courtesy of Fabrizio

Web Mining Web Mining Web mining is the use of data mining techniques to automatically

DATA MINING LECTURE 1 Introduction What is data mining? After years of data mining there is

Distributed frequent sequence mining with declarative subsequence constraints Alexander

THE DATA MINING PIPELINE What is data? The data mining pipeline: collection, preprocessing,

Introduction What is data mining? to Data Mining: On what kind of data? Data Mining

LECTURE 1: INTRODUCTION TO DATA MINING Dr. Dhaval Patel CSE, IIT-Roorkee What is data mining?

The DESQ Framework for Declarative and Scalable Frequent Sequence Mining Kaustubh Beedkar 1 Rainer

Data Mining: A Powerful Data Mining: A Powerful Tool for Data Cleaning Tool for Data Cleaning

Modeling Data the different views on Data Mining Views on Data Mining Fitting the data

What is SQL? Declarative Say what to do rather than how to do it Introduction

Data Mining: Concepts and Techniques Web Mining Li Xiong Slides credits: Jiawei Han and

Introduction to Web Mining What is Web Mining? Discovering useful information from the

Declarative Visitors to Ease Fine-grained Source Code Mining with Full History on Billions of

Exploration of declarative languages applicability to development of large-scale data processing

Declarative MapReduce 1 Declarative Languages Describe what you want to do not how to do it The

Contents Text Mining Concept Tasks Twitter Data Analysis with R Twitter Extracting Tweets

Introduction to Data Mining Methods and Tools by Michael Hahsler Agenda What is Data Mining?

SQL data manipula.on language SQL Data Manipulation Language (DML) Primarily declarative

Programming Language Ideas Escape the Lab: A Declarative Data Description Language Kathleen

Week 5 Video 2 Relationship Mining Causal Mining Causal Data Mining These slides developed in