When is a Clone not a Clone? (and vice-versa) Contextualized - - PowerPoint PPT Presentation

School of Computing Kingston, Canada

Contextualized Analysis of Web Services

James R. Cordy David B. Skillicorn Douglas Martin Scott Grant

When is a Clone not a Clone?

(and vice-versa)

Motivation

— The Personal Web

— Rapidly growing number of web services makes it

increasingly difficult to find and choose the right ones

— Need a quick and convenient

way to find alternatives

— Hand tagging impractical –

automation is needed!

— Automation

— Similarity detection techniques offer solutions! — Code clone detection from

software engineering research can find similar code fragments – why not similar services?

— Topic models from data mining

research can find text documents with similar semantics – why not similar services?

Motivation

Web Service Similarity

— Web services are stored in

service registries, containing WSDL service description files

— Could apply clone detection to

entire service descriptions

— But what we really want are

similar service operations

<operation name="GetStock" > <input message="tns:GetStockRequest" /> <output message="tns:GetStockResponse" /> </operation> <operation name="GetStock" > <input message="tns:GetStockRequest" /> <output message="tns:GetStockResponse" /> </operation>

<complexType name=“Stock”> <sequence> <element name=“Supplier” type=“xsd:string”/> <element name=“Warehouse” type=“xsd:string”/> <element name=“OnHand” type=“xsd:string”/> <element name=“OnOrder” type=“xsd:string”/> <element name=“Demand” type=“xsd:string”/> </sequence> </complexType >

Let’s try it!

<complexType name=“Stock”> <sequence> <element name=“date” type=“xsd:string”/> <element name=“open” type=“xsd:float”/> <element name=“high” type=“xsd:float”/> <element name=“low” type=“xsd:float”/> <element name=“close” type=“xsd:float”/> <element name=“volume” type=“xsd:float”/> </sequence> </complexType >

<operation name=“DrawRateChartCustom”> <input message=“DrawRateChartCustomIn”/> <output message=“DrawRateChartCustomOut”/> </operation> <operation name="GetTopicBinaryChartCustom"> <input message="GetTopicBinaryChartCustomSoapIn"/> <output message="GetTopicBinaryChartCustomSoapOut"/> </operation>

How about these?

So what went wrong?

— At this point we thought maybe our idea wasn’t

going to work — Maybe clone detection can’t help with

web service discovery?

— But why? What’s so special about WSDL?

Web Service Description Language (WSDL)

— A WSDL service description has

3 main parts:

Web Service Description Language (WSDL)

— A WSDL service description has

3 main parts:

— a <portType> element where the

• perations are declared;

Web Service Description Language (WSDL)

— A WSDL service description has

3 main parts:

— a <portType> element where the

• perations are declared;

— <message> elements

corresponding to inputs, outputs and faults of the operations;

Web Service Description Language (WSDL)

— A WSDL service description has

3 main parts:

— a <portType> element where the

• perations are declared;

— <message> elements

corresponding to inputs, outputs and faults of the operations;

— and a <types> element

containing an XML Schema that defines the data and structure types used in the messages

Web Service Description Language (WSDL)

— This simple example service

has two operations:

Web Service Description Language (WSDL)

— This simple example service

has two operations:

— ReserveRoom

Web Service Description Language (WSDL)

— This simple example service

has two operations:

— ReserveRoom — GetAvailableRooms

Web Service Description Language (WSDL)

— WSDL service description files contain descriptions

• f the operations that a web service has to offer

— But the pieces of each operation’s own description

are scattered over different parts of the WSDL file

— Difficult to identify complete units to analyze and

compare

The Problem

— This poses a problem for analysis techniques:

— Operations cannot easily be compared for similarity

using clone detectors, because there are no contiguous fragments to compare

— And they cannot be analyzed using data mining topic

models, because there are no separate complete documents to generate a model from

Our Solution

— Our solution is to contextualize the original

<operation> elements, to create self-contained

• peration descriptions

— We use source transformation to inline remote

information from the context into the elements that reference or depend on them

— We call these contextualized WSDL operations

Web Service Cells, or WSCells — The first example of a new kind of clone detection:

contextual clones

Contextualizing WSDL Operations

Contextual Clone Detection

An Experiment

— We have run an experiment to investigate the

difference between clone detection on WSCells and original raw operations — Two sets of WSDL service description files:

1,100 operations and 7,500 operations

— Compared NICAD clone detector results for each

set at various near-miss difference thresholds

0% = exact clone,

10% = 1 line in 10 different, and so on

An Experiment

— Number of clones decreases with WSCells

Difference ¡ Threshold ¡ Clone ¡Pairs ¡in ¡Set ¡1 ¡ Clone ¡Pairs ¡in ¡Set ¡2 ¡ Originals ¡ WSCells ¡ Originals ¡ WSCells ¡ 0.0 ¡ 852 ¡ 705 ¡ 1434 ¡ 1066 ¡ 0.1 ¡ 852 ¡ 734 ¡ 1434 ¡ 1228 ¡ 0.2 ¡ 879 ¡ 775 ¡ 1438 ¡ 1637 ¡ 0.3 ¡ 884 ¡ 813 ¡ 1469 ¡ 1637 ¡

<operation name="GetStock" > <input message="tns:GetStockRequest" /> <output message="tns:GetStockResponse" /> </operation> <operation name="GetStock" > <input message="tns:GetStockRequest" /> <output message="tns:GetStockResponse" /> </operation> <complexType name=“Stock”> <sequence> <element name=“Supplier” type=“xsd:string”/> <element name=“Warehouse” type=“xsd:string”/> <element name=“OnHand” type=“xsd:string”/> <element name=“OnOrder” type=“xsd:string”/> <element name=“Demand” type=“xsd:string”/> </sequence> </complexType > <complexType name=“Stock”> <sequence> <element name=“date” type=“xsd:string”/> <element name=“open” type=“xsd:float”/> <element name=“high” type=“xsd:float”/> <element name=“low” type=“xsd:float”/> <element name=“close” type=“xsd:float”/> <element name=“volume” type=“xsd:float”/> </sequence> </complexType >

— Reduction in

false positives

— Number of clone classes can increase with WSCells

Difference ¡ Threshold ¡ Clone ¡Classes ¡in ¡Set ¡1 ¡ Clone ¡Classes ¡in ¡Set ¡2 ¡ Originals ¡ WSCells ¡ Originals ¡ WSCells ¡ 0.0 ¡ 169 ¡ 187 ¡ 587 ¡ 433 ¡ 0.1 ¡ 169 ¡ 139 ¡ 587 ¡ 499 ¡ 0.2 ¡ 172 ¡ 142 ¡ 589 ¡ 631 ¡ 0.3 ¡ 171 ¡ 136 ¡ 591 ¡ 631 ¡

An Experiment

<operation name="GetStock" > <input message="tns:GetStockRequest" /> <output message="tns:GetStockResponse" /> </operation> <operation name="GetStock" > <input message="tns:GetStockRequest" /> <output message="tns:GetStockResponse" /> </operation> <complexType name=“Stock”> <sequence> <element name=“Supplier” type=“xsd:string”/> <element name=“Warehouse” type=“xsd:string”/> <element name=“OnHand” type=“xsd:string”/> <element name=“OnOrder” type=“xsd:string”/> <element name=“Demand” type=“xsd:string”/> </sequence> </complexType > <complexType name=“Stock”> <sequence> <element name=“date” type=“xsd:string”/> <element name=“open” type=“xsd:float”/> <element name=“high” type=“xsd:float”/> <element name=“low” type=“xsd:float”/> <element name=“close” type=“xsd:float”/> <element name=“volume” type=“xsd:float”/> </sequence> </complexType >

— Splits by deeper

differences – more precision

Clone Detection for Web Services

— Contextual clone detection with WSCells works! — Not only finds similar web service operations,

but uncovers similar operations we could not find in any other way

<operation name=“DrawRateChartCustom”> <input message=“DrawRateChartCustomIn”/> <output message=“DrawRateChartCustomOut”/> </operation> <operation name="GetRealChartCustom"> <input message="GetRealChartCustomSoapIn"/> <output message="GetRealChartCustomSoapOut"/> </operation> <operation name="GetLastSaleChartCustom"> <input message="GetLastSaleChartCustomSoapIn"/> <output message="GetLastSaleChartCustomSoapOut"/> </operation> <operation name=“DrawYieldCurveCustom”> <input message=“DrawYieldCurveCustomIn”/> <output message=“DrawYieldCurveCustomOut”/> </operation> <operation name="GetTopicChartCustom"> <input message="GetTopicChartCustomSoapIn" /> <output message="GetTopicChartCustomSoapOut" /> </operation> <operation name="GetTopicBinaryChartCustom"> <input message="GetTopicBinaryChartCustomSoapIn"/> <output message="GetTopicBinaryChartCustomSoapOut"/> </operation>

Semantic Analysis of Web Services

— Contextualized WSCells also make it possible to use

data mining topic models to do semantic analysis

• f web services

— Because they provide self-contained documents of

significant size

— Might topic models provide a different view

• f web service similarity?

Latent Dirichlet Allocation

— Latent Dirichlet Allocation (LDA) :

— A statistical model to uncover latent topics — Identifies the correlation between documents in

terms of shared latent topics (sets of tokens)

— Accepts a set of documents (e.g., source files) as

input, returns probability distributions over inferred topics (a topic model) as output — Each document has some probability of being related

to topic 1, another probability for topic 2, and so on

— Similar documents should be related to similar topics

Latent Dirichlet Allocation

— Documents are represented in the model in terms

• f probability distributions over topics

— Similarity between documents is found using the

Hellinger Distance — A measure of how much agreement there is between

the shared topics of two documents

— Almost identical documents have a small Hellinger

Distance since they will be related to the same topics

— In terms of web services, small Hellinger Distances

indicate highly related operations

Evaluating WSCells

— To evaluate the use of WSCells with LDA, we :

— Generate an LDA model for the original <operation>

elements, and another for the contextualized WSCells

— Explore the Global and Local Similarity between each

pair of operations in the models

— Global Similarity an overall view of the most closely

related web service operations in the service set

— Local Similarity a per-operation view of the other

most related web service operations for each

• peration

Global Similarity

— We look at Global Similarity using a visualization

called Bluevis

— Bluevis shows the global conceptual structure of a

system by highlighting similar operations using an illuminated line from left-to-right — Plot some top fraction of similar operations

(top 25,000 in our examples)

— Use a consistently ordered list of web service

• perations for the LDA model to view the differences

— If a display is noisy, it is often an indication that the

model is not identifying meaningful data

Global Similarity

Global Similarity

— For original raw operations:

— Bluevis highlights the LDA

most similar operations

— Some clear structure — However, most of this is

due to shared keywords, like get and SOAP

— This uncontextualized

model has very little value

Global Similarity

Global Similarity

— For contextualized WSCells:

— A clearer semantic

structure, less noise overall

— Operation similarity

becomes meaningful

— Services with semantic

similarity discovered — E.g., Operations with

similar parameters or faults, such as those that manipulate holiday dates

• r financial rates

Local Similarity

— We can also examine the local similarity for each

individual operation — Identify the complete ordered list of similarity scores

for an operation in the data set

— Using the top similarity scores, evaluate how

meaningful the data is from a user's perspective — For example, how can I find the most similar web

service operations to the one I am using now?

— We use a tool called POCO (Pairwise Observation of

Concepts) to examine the most similar operations

Local Similarity

Local Similarity

Operation Most similar WSCell Most similar original raw WSDL operation ListFinancials GetFinancialServicesFromList LanguagesList ExportShipsAndCategories ExportIteneraryAndSteps Search GetIssueData GetFlightData word_cloud GetWeatherReport GetWeather GetIndices GetAIDIBOR GetTRLIBOR GetCarriers searchByIdentifier searchByNameAndAddress GetLastSecurityHeadlines ToolsAndHardwareBox KitchenAndHousewareBox ListRenditions GetReservations GetRoomAvailabilityForDay GetSOFIBOR GetOtherProductInfo NextOtherProductPortion GetParkingInfo GetAllSplitsByExchange GetAllCashDividendsByExchange GetTeamLoyalties2

Summary

— Very-high-level domain-specific languages such as

WSDL make poor targets for similarity analysis using clone detection and topic models — Lack of local context prevents meaningful results

— Contextualizing using WSCells exposes both cloning

and semantic relationships between web operations — Clone detection of WSCells identifies similar web

service operations

— Topic models of WSCells expose both global

system-wide semantic relationships and local individual relationships between operations

Current & Future

— Continue analysis of web services for the Personal

Web using our results

— Apply contextualization to similarity analysis of

• ther modeling and specification languages

(currently Simulink, Stateflow and UML sequence diagrams)

— Experiment with effect of contextualization on

clone and topic model analysis of traditional languages such as Java and C (“contextual clones”)

James R. Cordy David B. Skillicorn Douglas Martin Scott Grant

Questions?

Contextualized Analysis of Web Services

When is a Clone not a Clone?

(and vice-versa)