A Review -Dynamic Analysis of Web System by using Model-Based Testing and Process Crawler Model

DOI : 10.17577/IJERTV6IS060512

Download Full-Text PDF Cite this Publication

  • Open Access
  • Total Downloads : 128
  • Authors : Nayan Mulla, Prof. Sachin B. Takmare, Prof. Pramod A. Kharade
  • Paper ID : IJERTV6IS060512
  • Volume & Issue : Volume 06, Issue 06 (June 2017)
  • DOI : http://dx.doi.org/10.17577/IJERTV6IS060512
  • Published (First Online): 01-07-2017
  • ISSN (Online) : 2278-0181
  • Publisher Name : IJERT
  • License: Creative Commons License This work is licensed under a Creative Commons Attribution 4.0 International License

Text Only Version

A Review -Dynamic Analysis of Web System by using Model-Based Testing and Process Crawler Model

Mrs. Nayan Mulla1, 1Department of Computer Science and Engg., Bharati Vidyapeeths College of Engineering,

Kolhapur, Maharashtra, India

Prof. Sachin B. Takmare2, 2Department of Computer Science and Engg., Bharati Vidyapeeths College of Engineering,

Kolhapur, Maharashtra, India

Prof. Pramod A. Kharade3 3Department of Computer Science and Engg., Bharati Vidyapeeths College of Engineering,

Kolhapur, Maharashtra, India

Abstract: Modern business applications predominantly rely on web technology, enabling software vendors to efficiently provide them as a service, removing some of the complexity of the traditional release and update process. To increasing web application accuracy and speed user process crawler model.Cutting edge business applications transcendently depend on web innovation, empowering programming sellers to give proficiently them as an administration, uprooting a portion of the multifaceted nature of the customary discharge and overhaul process. While this encourages shorter, more productive and successive discharge cycles, it obliges persistent testing. Having knowledge into application conduct through unequivocal models can to a great extent bolster improvement, testing and support. Model-based testing permits effective test creation taking into account a depiction of the states the application can be in and the moves between these states. As determining conduct models that are sufficiently exact to be executable by a test computerization device is a hard assignment, an option is to concentrate them from running applications.

Keywords: Specification Mining; Dynamic Analysis; Model-based Testing; Web System Testing.


    Then again, mining such models is a test, specifically in light of the fact that one needs to know when two states are proportional, and also how to achieve that state. Here introduce ProCrawl (Process Crawler), a device to mine conduct models from web applications that backing multi- client work processes. ProCrawl incrementally takes in a model by creating system runs and watching the application conduct through the client interface. In our assessment on a few true web applications, ProCrawl removed models that briefly depict the actualized work processes and can be specifically utilized for model-based testing[14].

    into a prevailing customer for big business programming. Also, the accessibility of system data transfer capacity empowers applications to be worked by the merchant and gave as administrations to clients. Working applications on the seller side evacuates a portion of the intricacy and expenses of the customary programming discharge and redesign procedure; while this empowers shorter, more proficient and incessant discharge cycles with a littler number of components, it puts more weight on programming advancement and obliges giving careful consideration to operational viewpoints, nonstop Quality Affirmation (QA) and testing. Having knowledge into the conduct of a product segment through unequivocal models can to a great extent enhance the improvement, QA and support process.

    The field of determination mining so as to mine plans to encourage these exercises deliberations from projects and their executions; commonly, models of the program's conduct.

    On the off chance that these models are sufficiently exact, they can even be utilized as post-facto details of the project and test designers can apply them in a ceaseless joining environment to check for relapses after code changes. Determination mining has been utilized to infer effectively aphoristic details, for example, capacity and information invariants from projects or limited state machines portraying states and moves for individual classes For such little scale spaces, it is genuinely simple to approve details, in light of the fact that both system code and project state are available and agreeable to typical thinking and thorough testing. Extricating models on framework level is a great deal more troublesome. Project code and system state, for the case, may not be accessible for examination, as the application may be appropriated over a few layers and locales. As a rule, the main suspicion that can be made is that there is some client interface (UI, for example, a web front end that takes into consideration human communication.)


    The Integrating existing tests with Numerous Web applications accompany existing unit and framework tests. Investigating intends to incorporate and adjust these tests into computerized slithering.

    Research development challenges – utilizing their information for data provisioning, and their association streams for shockingly better scope. The base of our model is absolutel state-based. Here considering utilizing setting free and connection delicate language structures that would permit to express a great deal more perplexing associations and conditions. Besides Web applications, the strategies connected on nonspecific GUI-driven applications, giving model extraction and resulting model-construct testing in light of an extensive variety of stages and projects.

    Overview of solution-

    Richer models- At this point, the base of our model is purely state-based. We are thinking about leveraging context-free and context-sensitive grammars, that would allow to express much more complex interactions and dependencies.

    Alternative platforms- Besides Web applications, the PROCRAWL techniques could just as well be applied on generic GUI-driven applications, providing model extraction and subsequent model-based testing on a wide range of platforms and programs


    Conduct models can emotionally supportive network comprehension and acceptance. Model-Based Testing (MBT) takes into account productive test creation when a model depicting the conceivable and the normal application conduct is accessible, and along these lines for expanded computerization. On the other hand, physically making and keeping up conduct models that are sufficiently exact to be executable by a test computerization apparatus all through the product improvement procedure is costly. Web applications commonly come without express models, which suggests for the most part manual and in this manner less proficient test creation, furthermore eases off comprehension and support[1]. There exist many approaches for reverse engineering various properties of software components, PROCRAWL is a dynamic technique mining behavior models of web applications that generalize upon multiple runs (traces) of the application, which makes it applicable where source code is not accessible or amenable to static analysis. The resulting models capture the interaction of multiple users, as well as the interplay between input data and event sequencing affecting

    application behavior.

    PROCRAWL is related to dynamic techniques mining (extended) finite state machines, experimental techniques that systematically generate program runs to learn program behavior, techniques extracting models from GUI applications, and web crawling approaches[2].

    There are various approaches that mine FSMs capturing program behavior as sequencing of events, many of them building upon the k-tails algorithm However, program behavior usually depends not only on the sequencing of events, but also on the provided input data, which led to

    approaches combining FSM inference with data rule inference[3].

    MINER mines parametric specifications capturing the multi-object behavior of Java classe by preprocessing parametric execution traces and using an extension of k-tails to extract an FSM with data parameters. Pradel et al[4]. Also mine parametric multi-object specifications (API usage protocols) from Java classes. While multi-object specifications capture method calls performed on multiple objects, PROCRAWL captures UI interactions performed by multiple actors. One pioneering work was developed by Lorenzo et al. While nondeterministic models may be acceptable for the program comprehension, they are usually too imprecise for test case generation, the transition guard learning approach implemented in PROCRAWL is similar to the approach of Walkinshaw et al[5].

    The quality of models extracted by dynamic techniques closely depends on the

    choice of program runs (traces) these techniques generalize upon. Experimental approaches such as PROCRAWL tackle this problem by systematically generating runs to explore program behavior; this allows PROCRAWL to verify its hypotheses and explicitly control the exploration scope, which would otherwise be implicitly induced by a given set of traces. Such techniques have successfully been implemented for mining behavior models of Java classes[6].

    The underlying problem PROCRAWL has to solve is an instance of online exploration of a directed multi-graph by repeatingly selecting an outgoing edge from the current vertex and traversing it, which is a fundamental problem in robotics and has been extensively studied for strongly connected graph However, ABMs are usually not strongly connected, i.e. PROCRAWL might need to reset the SUT to the initial state to continue graph exploration[7].

    PROCRAWL mines program behavior by observing changes on the applications web UI, which is related to GUI Ripping developed by Atif Memon et al. GUITAR reverse engineers Event-Flow Graphs (EFGs,) of Java desktop, the web and Android applications, which are used for model- based GUI testing. While in ABMs extracted by PROCRAWL user actions are modeled as transitions between nodes that represent abstract GUI states of multiple users, in EFGs they are modeled as nodes with transitions representing the event flow[8].

    To explore a web application, PROCRAWL applies techniques similar to web crawling. Gives an overview on the state of the art. A prominent representative and actively developed crawler for AJAX applications is CRAWLJAX , which automatically creates a State-Flow Graph (SFG) of the dynamic DOM states and the event-based transitions between them. SFGs depict the various navigational paths and UI states within an AJAX application. Although ABMs look similar to SFGs, the underlying abstraction in SFGs is much closer to the SUTs UI. Limiting the state abstraction scope to a single DOM tree of a single user, limits the number of actions that can be detected by the crawler and often leads to a nondeterministic FSM, prohibiting effective model-based testing. In ABMs, a node represents an abstraction over

    multiple DOM trees (views) of multiple users and transitions refer to generated scripts encapsulating sequences of UI commands. Furthermore nondeterminism is effectively eliminated by learning transition guards over the input data[9][10].

    In our evaluation we apply the behavior models inferred by PROCRAWL for state-based web application testing, similar to and However, due to the wider exploration scope and higher level of abstraction supporting multiple users and views, the test cases generated from PROCRAWL ABMs are more than pure UI tests and suitable for testing workflows. reports the state of the art in web testing[5][11]. Presents an approach for reverse engineering business processes exposed in web applications by inferring an FSM from execution traces and transforming it to Business Process(BPMN) and Recovering and Reducing Business Processes REBPMN [12]

  4. PROBLEM STATEMENT Implementation of integrating the system with automated process crawling model, using their data for input provisioning, and their interaction flows for even better coverage

  1. Design an Web applications GUI-driven applications providing model extraction and subsequent model-based testing on a wide range of platforms and programs.

  2. To integrate and adapt these tests into automated crawling, using their data for input provisioning, and their interaction flows for even better coverage.

      1. Methodology

        1. UI Command Abstraction


      1. Proposed System Architecture

        Figure 1. ProCrawl overview

        ProCrawl applies a black-box approach that is it mines application behavior without accessing the sources code of the System Under Test(SUT).It determines the configured set of UI views(technical DOM trees),

      2. Scope

        To fully automatic configurable tool to my workflow models from web applications. It includes implements an iterative approach of executing actions through the UI, observing changes to the UI, and enhancing the model. Design an approach to increases model accuracy by inferring transition guard conditions from the input data.

      3. Objective of proposed work

    1. To Study and Integrating existing tests and implementation GUI-driven system using proCrawl model with context-free grammar and context sensitive grammar.

    2. To Study context-free and context-sensitive grammars, that would allow to express much more complex interactions and dependencies besides

    Figure 2.First iteration. PROCRAWL

    It executes an action with a congured actor, infers the state of the SUT by applying an abstraction function over the DOM trees (views) extracted with the actors, and update the behavior, model.

      1. Modules

        To integrate our system described following four models

        1. Model 1: Configuration

          Configured system with the identified actors, views, start action, a test fixture, and the click depth and started system and without transition guard learning;

        2. Model 2: Automatic exploration

          Based on this configuration, the system automatically explores the behavior of the web application.

          each iteration consists of the following steps 1 Run action.

          1. State abstraction.

          2. Model.

        3. Model 3: Application Behavior Model

          By generating tests from the ABM of the baseline application and executing these tests on the delta build, we are able to detect

          • Logical Changes (LC) to the underlying workflow, e.g. removing workflow steps or introducing additional mandatory steps.

          • Structural Changes (SC) to the web UI that cause the generated scripts to fail, e.g. removing, replacing or changing attributes of UI elements that are used in the scripts to locate these elements.

        4. Model 4: Model Based Testing and validation

          With the aim of increasing test automation, Model-Based Testing (MBT) has expanded the automation of software testing towards the test design phase: behavior models can be used to derive a test suite that is when executed through a test automation framework. The Industrial application has positive effects of MBT on the development productivity .However, wide-scale adoption suffers from the absence of explicit models that are complete enough to be executable by test frameworks.

      2. Algorithm

        1. Algorithm 1 shows the initialization of the exploration procedure for mining the behavior model [6]

          Algorithm 1: MAIN Input: cong

          1. global A new HashMultimap();

          2. global ABMk new AppBehaviorModel(); 3 global driver INITDRIVERPOOL(cong);


            12 A(s0) A(s0){(,)}; 13 end

            1. end

            2. t (s,action,s0);

            3. ABMk.trans ABMk.trans{t};

            4. PLUGINS.TRANSITION(t);

            5. else if CLICKS() < k then 19 foreach e E do

            20 e h + (click,le,null)i; 21 A(s) A(s){(,)}; 22 end

            1. end

            2. EXPLORE(s0);

            3. else if {s ABMk.S | A(s) 6= }6=

            4. sp GOTOPENDINGSTATE(s);

            5. EXPLORE(sp);

            6. end


4 s0 DETERMINESTATE(); 5ABMk.initialState s0;

6A(s0) {cong.startAction};



    1. EXPLORE(s0);


      1. Algorithm 2. Set of elements extracted from the DOM trees of multiple actor/view relations[6] Algorithm 2: DETERMINESTATE

        Data: cong

        Output: current state s of the SUT 1s {};

        1. foreach (,v) cong.exploration_scope do

        2. DOM driver.GETDOM(,v);

        3. fDOM FILTER(, DOM);

        4. foreach e SELECT(,fDOM) do 6 s sf(,v,e);

        1. end

        2. end

        3. return s;

      2. Algorithm 3 shows the exploration procedure recursively building up the behaviour model[6]

Algorithm 3: EXPLORE Input: current state s of the SUT

  1. if A(s) 6= then

  2. action (,) A(s);

  3. A(s) A(s)\{action};

  4. E driver.EXECUTE(action); 5 s0 DETERMINESTATE();

  1. if s s then

  2. if s ABMk.S then

  3. ABMk.S ABMk.S{s0};


  5. foreach (,v,) s0\s0 do 11 hv + (click,l,null)i;

Evaluation of the designed system is on several real-world web applications are on the basis of adequate in size, accuracy, cover all to almost all workflow-relevant actions, and are a suitable input to model-based test generation. And using proCrawl model increasing performance and accuracy of the any web application.


  1. M. Utting and B. Legeard. Practical Model-Based Testing: A Tools Approach. Morgan Kaufmann Publishers Inc., San Francisco, CA,

    USA, 2007

  2. M. P. Robillard, E. Bodden, D. Kawrykow, M. Mezini, and T. Ratchford. Automated API property inference techniques. IEEE Trans. Softw. Eng., 39(5):613637, May 2013.

  3. N. Walkinshaw, B. Lambeau, C. Damas, K. Bogdanov, and P. Dupont. Stamina: A competition to encourage the development and assessment of software model inference techniques. Empirical Softw. Eng., 18(4):791824, Aug. 2013..

  4. C. Lee, F. Chen, and G. Ros¸u. Mining parametric specications. In ICSE, pages 591600, New York, USA, 2011.

  5. M. Pradel, C. Jaspan, J. Aldrich, and T. R. Gross. Statically checking api protocol conformance with mined multi-object specications. In ICSE, pages 925935, Piscataway, NJ, USA, 2012. IEEE Press.

  6. M. Schur, A. Roth, and A. Zeller. Procrawl: mining test models from multi-user web applications. In International Symposium on Software Testing and Analysis, ISSTA 14, San Jose, CA, USA July 21 – 26, 2014, pages 413416, 2014.

  7. V. Dallmeier, N. Knopp, C. Mallon, S. Hack, and A. Zeller. Generating test cases for specication mining. In ISSTA, pages 8596, New York,

    USA, 2010. ACM

  8. B. N. Nguyen, B. Robbins, I. Banerjee, and A. Memon. GUITAR: an innovative tool for automated testing of GUI-driven software.

    Automated Software Engineering, pages 141, 2013

  9. W. Grieskamp, N. Kicillof, K. Stobie, and V. Braberman. Modelbased quality assurance of protocol documentation: tools and methodology. Software Testing, Verification & Reliability, 21(1):5571, Mar. 2011.

  10. V. Garousi, A. Mesbah, A. Betin-Can, and S. Mirshokraie. A Systematic Mapping Study of Web Application Testing. Information and Software Technology, 55(8):13741396, Aug. 2015

  11. M. Schur, A. Roth, and A. Zeller. Mining behavior models from enterprise web applications. In ESEC/SIGSOFT FSE, pages 422432, 2013.

  12. A. Tomasi, A. Marchetto, C. D. Francescomarino, and A. Susi. reBPMN : Recovering and Reducing Business Processes. In Software Maintenance (ICSM), 2012 28th IEEE International Conference on, pages 666669. IEEE, 2012

Author Profile

Mrs. Nayan Mulla is a PHP developer at Purestudy Software Services Pvt..Ltd and a M.E student at Shivaji University. She received the B.E. degree in Information Technology in 2013 from Shivaji University. His research concerns the analysis and testing of enterprise web application.

Mr. Takmare Sachin Balawant is working as Associate Professor and Head of Department in computer Science and Engineering Department of Bharati Vidyapeeths college of Engineering, Kolhapur with teaching experience of about 10 year. He has published bout three International Papers and five National Papers.

Mr. Pramod A. Kharade is working as Assistant Professor of Department in computer Science and Engineering Department Bharati Vidyapeeths college of Engineering, Kolhapur.

Leave a Reply