Project Proposals for 2016

Proposal 1: Native inbound and connector for IBM MQ 7.5

Description

Currently with WSO2 Enterprise Service Bus recommended way to connect to IBM MQ is using JMS transport. But this has limitations and sometimes gives issues when integrating. 

Deliverables

• Custom inbound endpoint as consumer.
• Connector as producer.

Skills Needed

• JMS Transport
• Message brokers
• Java

References

[1] JMS Inbound Protocol

[2] Configure with IBM WebSphere MQ

[3] http://mrmalakasilva.blogspot.com/2013/10/connecting-mechanisms-other-than.html

[4] Creating a Third Party Connector and Publishing in WSO2 Connector Store

Possible Mentor/s

Malaka Silva

Sajini De Silva 

Proposal 2: Developer Tooling for Microservice Framework for Java

Description

Microservice Framework for Java (MSF4J) [1]  is a lightweight, fast runtime that can host annotation based services, such as JAX-RS in a micro service architecture[2]. Microservices developed using WSO2 MSF4j can start in just a few milliseconds in a Docker container and can easily be added to a Docker image definition.

 MSF4J requires development tooling to make it appealing to the Java Development Community. It requires end-to-end tooling that helps developer to implement a microservice, run and unit-test a microservice in the integrated development environment itself.

Deliverables

• IntelliJ IDEA Plugin for running a Java Service in MSF4J

• Eclipse Plugin for running a Java Service in MSF4J

Skills Needed

• Java

• Eclipse plugin development

• IntelliJ plugin development

References

[1] http://wso2.com/products/microservices-framework-for-java/

[2] http://martinfowler.com/articles/microservices.html

Possible Mentor/s

Dimuthu Leelarathne

Proposal 3: [ML] Visual (Deep) Neural Network Builder

Unlike other machine learning algorithms, neural networks can be constructed using few basic computational units such as logistic and rectifier linear units (ReLU). Depending on the number of such units per layer and number of layers in the network, it is possible to create a wide range of neural network architectures.

The main idea of this project is to build a visual interface that helps people to easily create and train neural networks by just dragging and dropping basic computational units. The main workflow of the visual builder would be:

1. Create an input layer and training/testing dataset will be fed to the input layer.
2. Next,  create one or more hidden layers and specify connections between input/hidden and hidden/hidden layers. Also, he needs to specify the computational units use in the hidden layers.
3. Next step would be, specifying the output layer. Depending on the problem (whether it is classification or regression), the user might pick a suitable output layer from available layers.

4. In addition to above, other parameters such as learning rate, regularization and etc. should be able to configure inside the visual editor.

5. Finally, it should be possible to train the network using visual network builder.

Nice to have features: hyper-parameter optimization using cross-validation, graphs for visualizing training/cross-validation errors. 

Deliverables

Visual neural network builder integrated with WSO2 Machine Learner.

Skills Needed

• Basic Machine Learning knowledge.
• Java/Apache Spark/DeepLearning4J
• Jaggery/JavaScript/HTML

References

[1]. http://deeplearning4j.org/
[2]. http://spark.apache.org/

Possible Mentor/s

Upul Bandara ( upul AT wso2 DOT com )

Nirmal Fernando ( nirmal AT wso2 DOT com)

Proposal 4: [ML] Ensemble Methods Support for WSO2 Machine Learner

Description

In Machine Learning, Ensemble learning is a combining of multiple learning algorithms/models to obtain better predictive accuracy than what could be obtained from a single learning algorithm. Final prediction is done by taking a weighted vote of the predictions of the combined algorithms. Currently WSO2 machine learner does not facilitate this. The main tasks of the project is to:

• Implement a ensemble method(s), to combine multiple algorithms.
• Create a UI to include the end to end flow of training the algorithm.
• Integrate it to the WSO2 Machine Learner.
  
  

With the implemented method, user should be able to build an ensemble solution:

• Using models that are already trained.
• By training all the components (algorithms associated) of the ensemble method, as a set of completely new algorithms.

Deliverables

• Ensemble Method integration for WSO2 Machine Learner.

Skills Needed

• Basic Machine Learning knowledge
• Java
• Knowledge in HTML, js
• UI/UX skills

References

[1] https://docs.wso2.com/display/ML110/About+Machine+Learner

[2] https://github.com/wso2/product-ml

[3] https://github.com/wso2/carbon-ml

Possible Mentor/s

Supun Sethunga (supuns AT wso2 DOT com)

Nirmal Fernando (nirmal AT wso2 DOT com)

Proposal 5: [ML] Natural Language Processing Toolbox for WSO2 Machine Learner

Description

Currently, there are few natural language processing (NLP) functionalists implemented for WSO2 Machine Learner (ML) as independent components. But we need to have a unified end-to-end NLP solution which aligns with current WSO2 ML workflow. The main task of this project would be unifying existing NLP tools into one solution with possible extensions and relevant UI elements.

Deliverables

• Natural Language Processing Toolbox for WSO2 Machine Learner

Skills Needed

• Basic knowledge in machine learning
• Basic knowledge in natural language processing
• Java
• HTML, CSS, JS
• UI/UX skills
• Familiarity with ML/NLP libraries (Apache Spark, Deeplearning4j, Apache OpenNLP)

References

[1] https://docs.wso2.com/display/ML110/About+Machine+Learner

[2] https://github.com/wso2/product-ml

[3] https://github.com/wso2/carbon-ml

[4] http://spark.apache.org/

[5] http://deeplearning4j.org/

[6] https://opennlp.apache.org/

Possible Mentors

CD Athuraliya (chathurike AT wso2 DOT com)

Nirmal Fernando (nirmal AT wso2 DOT com)

Proposal 6: [ML] Predictive analytics with online data for WSO2 Machine Learner

Description

The idea of incremental learning with streaming data focuses on two objectives:

1. Identifying patterns in the recent history.
2. Updating the patterns with incoming data without catastrophic forgetting.

These objectives can be approached in two methods:

1. Incremental algorithms - there are machine learning algorithms which can be modified to support incremental learning. eg. mini-batch k-means, stochastic gradient descent based linear models, SVM, etc.
2. Periodic re-training - machine learning models are trained with buffered data periodically.

The student has to consider these points when designing the architecture of the system. Moreover, data horizon (how quickly a most recent data point becomes a part of the model) and data obsolescence (how long does it take a past data point to become irrelevant to the model) need to be taken into account and should be adjustable.

WSO2 ML uses Apache Spark for its' machine learning algorithms. Apache Spark already supports learning algorithms with streaming data such as k-means clustering and generalized linear models. Student may use these implementations in their systems. It is not required to support incremental learning for every algorithm available in the WSO2 ML as there are technical limitations in employing incremental learning with those.

Deliverables

• Designing an architecture for incremental learning and visualizations.
• Creating the incremental learning component.
• Creating interactive visualizations for incremental learning models.
• Allowing users to change parameters of the algorithms on the fly (during the incremental learning procedure) by analyzing the models. (Optional).
• Documentation and examples.
   

Skills Needed

• Basic knowledge in machine learning
• Basic knowledge about working with streaming data
• Java
• Scala (optional)
• UI/UX skills
• Familiarity with Apache Spark

References

[1] https://docs.wso2.com/display/ML110/About+Machine+Learner

[2] https://github.com/wso2/product-ml

[3] https://github.com/wso2/carbon-ml

[4] http://spark.apache.org/

[5] https://spark-summit.org/east-2015/talk/streaming-machine-learning-in-spark

Possible Mentors

Pruthuvi Maheshakya Wijewardena (maheshakya AT wso2 DOT com)

Nirmal Fernando (nirmal AT wso2 DOT com)

Proposal 7: [ESB/GW] Improve production system debugging capabilities of ESB (ESB Flight Recorder)

Description

When an issue occurs in a production server, to investigate the incident, we need to get several information from the server instance.
At the moment we need to extract these information manually and sometimes we may not get required information before the server goes to a catastrophic situation.
This project is about automatically extracting useful information from the server, when there are potential issues.

Following are the high level requirements
- Alerts on increase of resources
- Automatically generate required information for debugging (dumps)
- Handle/prevent catastrophic situations
- Correlate dumps with environment changes like request load and generate reports that can be used to analyse a problem

Deliverables

• ESB Production debugger component.

Skills Needed

• Java

References

https://docs.wso2.com/display/ESB490/WSO2+Enterprise+Service+Bus+Documentation

Possible Mentors

Isuru Udana (isuruu AT wso2.com)

Isuru Ranawaka (isurur AT wso2.com)

Proposal 8: [ESB/GW] HTTP Load balancer on top of WSO2 Gateway 

Description

Building a HTTP load balancer on top of WSO2 Gateway

Following are the high level requirements

• HTTP/S static load balancing 

• Fully compliant with HTTP specification  

• Performance : Must be able to compete with existing load balancers 

• Session persistence, Heartbeat/Health check, Redirects 

Deliverables

• HTTP Load balancer product milestones. 

Skills Needed

• Java, knowledge on Load balancing techniques 

References

[1] https://github.com/wso2/product-gw

Possible Mentors

Isuru Ranawaka (isurur AT wso2.com)

Kasun Indrasiri (kasun AT wso2.com)

Proposal 9: [ESB/GW] File Gateway on top of WSO2 Gateway 

Description

Building a File Gateway on top of WSO2 Gateway

Following are the high level requirements

• Managed file transfer   
• FTP, SFTP and other types of file transfer
• Transferring large files 
• Must cover all the file based integration user cases support in ESB. 

Deliverables

• File Gateway product milestones. 

Skills Needed

• Java  

References

[1] https://github.com/wso2/product-gw

[2] https://commons.apache.org/proper/commons-vfs/filesystems.html

Possible Mentors

Viraj Senevirathne (virajs AT wso2.com)

Kasun Indrasiri (kasun AT wso2.com)

Proposal 10: [ESB/GW] Security Gateway on top of WSO2 Gateway 

Description

Building a Security Gateway on top of WSO2 Gateway

Following are the high level requirements

• Securing your enterprise services from various attacks (XML threat protection, etc.)
• Controlling the inward traffic to your services (Filtering, Throttling)
• Provides security enforcement like OAuth2 to the services
• Analytics/Reporting of threats on the system

Deliverables

• Security Gateway product milestones. 

Skills Needed

• Java  

References

[1] https://github.com/wso2/product-gw

[2] http://www.websense.com/content/web-security-gateway-features.aspx

Possible Mentors

Chanaka Fernando (chanakaf AT wso2.com)

Prabath Siriwardana (prabath AT wso2.com)

Senduran Balasubramaniam (senduran AT wso2.com)

Proposal 11: [ESB/GW] WSO2 ESB mediation config migration tool

Description

Building a tool to migrate existing synapse configurations and data to a new configuration which is based on a new mediation language

Following are the high level requirements

• Building a model to map synapse configurations with the new mediation language
• Gracefully handle configurations which are not compatible across two different mediation languages
• Migrate data related to the server instances

Deliverables

• Migration tool milestones

Skills Needed

• Java 
• Domain Specific Languages (DSL)
• Synapse

References

[1] WSO2 Enterprise Service Bus Documentation

[2] http://martinfowler.com/books/dsl.html

Possible Mentors

Chanaka Fernando (chanakaf AT wso2.com)

Ravi Undupitiya (ravi AT wso2.com)

Proposal 13: [ESB/Tooling] Automated tests for ESB Graphical editor using SWTBot

Description

Currently we do not have automated tests for ESB graphical editor[1] which is a GMF based drag and drop tool to develop ESB configurations. We can use Eclipse SWTBot[2] to implement automated tests for these kind of GMF editors. Idea is to implement an automated test suite to cover graphical functionalities of ESB editor by composing mediation flows which involves Proxy services, APIs, Sequences, Endpoints, mediators and their properties etc.

Deliverables

• Automated test suite to test graphical functionalities of ESB graphical editor

Skills Needed

• Basic knowledge on Eclipse EMF/GMF
• Basic knowledge on WSO2 ESB configurations

• Java

References

[1] https://docs.wso2.com/display/DVS380/Creating+ESB+Artifacts

[2] http://www.eclipse.org/swtbot/

[3] https://wiki.eclipse.org/SWTBot/UsersGuide#GEF.2FGMF-based_editor_testing

[4] https://wiki.eclipse.org/SWTBot/Automate_test_execution 

Possible Mentors

Viraj Rajaguru (viraj AT wso2 DOT com)

Nuwan Pallewela (nuwanp AT wso2 DOT com)

Proposal 14: [ESB/Tooling] Support JSON, CSV, Java, Excel and Web services formats in Data Mapper tool.

Description

Today, enterprise applications use different data formats and different data structures. WSO2 Data Mapper tool can be used to map different data formats/structures between two systems/applications while system integrations. WSO2 Data mapper tool is based on Eclipse GMF and builds graphical input data model and output data model using provided Apache Avro[1] schemas.

Developers can extend Data mapper tool to support any data format by adding a data type transformer to generate Avro schema based on the relevant data format(Eg: JSON, CSV etc). This project is to write data format transformers to transform CSV, JSON, Java, Excel, Web services formats to Avro format. 

Deliverables

• Components for generating Apache Avro schemas for data formats such as JSON, Java, Excel, Web services.

Skills Needed

• Java
• Basic knowledge on Apache Avro

References

[1] https://avro.apache.org/docs/1.7.7/spec.html

Possible Mentors

Viraj Rajaguru (viraj AT wso2 DOT com)

Nuwan Pallewela (nuwanp AT wso2 DOT com)

Proposal 15: [ESB/Tooling] ESB Visual editor enhancements

Description

WSO2 Developer Studio provides comprehensive graphical composition tool to create ESB configurations[1] for WSO2 ESB. ESB Graphical editor is a drag and drop tool where users can develop, debug, deploy and test their configurations in a cost-effective manner. We have identified few areas in this tools which can be improved further for a better user experience.

Followings are the high level requirements

• Provide ability to aggregate properties in to a single element when we have large number of property mediators 
• Utilize editor space properly 
• Enhance editor for complex use cases
• Improve performance of editor to gain smoothness

Deliverables

• Improve ESB graphical editor in above mentioned areas.

Skills Needed

• Java
• Basic knowledge on Eclipse GMF

References

[1] https://docs.wso2.com/display/DVS380/Creating+ESB+Artifacts

Possible Mentors

Viraj Rajaguru (viraj AT wso2 DOT com)

Isuru Udana (isuruu AT wso2 DOT com)

Proposal 16: [DevStudio kernel/Tooling] Contact Development Team feature in DevStudio.

Description

Deliverables

• Eclipse plugin.

Skills Needed

• Java
• Eclipse plugin development 

References

[1] http://help.eclipse.org/juno/index.jsp?topic=%2Forg.eclipse.platform.doc.isv%2Freference%2Fextension-points%2Findex.html

Possible Mentors

Jasintha Dasanayaka (jasintha AT wso2 DOT com)

Kavith Lokuhewage (kavith AT wso2 DOT com)

Proposal 17: [CEP] Auto-tuning of WSO2 CEP

Description

Deliverables

• Auto-tuner for WSO2 CEP

Skills Needed

• Java
• Knowledge on WSO2 CEP

References

[1] http://dl.acm.org/citation.cfm?doid=2370816.2370826 

[2] http://dl.acm.org/citation.cfm?doid=2528412

Possible Mentors

Miyuru Dayarathna (miyurud AT wso2 DOT com)

Sriskandarajah Suhothayan (suho AT wso2 DOT com)

Proposal 18: [ESB] X12 inbound and connector

Description

x12 is a standard that is been used in health care industry. Currently with ESB only hl7 support is available. Based on the research need to provide solution to support x12 messages.

Deliverables

• x12 custom inbound and connector

• Documentation and guide

Skills Needed

• Java
• Knowledge on WSO2 ESB connectors and inbound
• Health care domain to some extend

References

[1] http://www.hl7.com/x12.html

Possible Mentors

Miyuru Dayarathna (miyurud AT wso2 DOT com)

Sriskandarajah Suhothayan (suho AT wso2 DOT com)

Proposal 19: [PC] Web (Javascript / D3) based BPMN editor to support a subset of commonly used BPMN constructs

Description

WSO2 Process Center (PC) is a product to facilitate business process governance at multiple levels from business users to technical users. A javascript based BPMN editor is required to allow PC users to define business processes using a graphical notation from the web UI. This BPMN editor focuses on supporting a subset of BPMN, which is sufficient to create abstract process models. Developing executable process models is not considered at this stage.

Main task is to create a javascript/D3 based graph editor to construct BPMN models. BPMN constructs like tasks, events and gateways have to supported. Users should be able to draw BPMN diagrams by dragging and dropping BPMN constructs from a pallet and positioning them in a canvas. Resizing, repositioning, deleting, etc of BPMN constructs have to be supported.

Editor has to be developed in an extensible manner so that it can easily used for constructing other (graph-based) diagrams such as flowcharts with minimal changes.

Deliverables

• Javascript/D3 based BPMN editor
• Documentation on extending the editor

Skills needed

• Strong javascript knowledge
• D3

References

http://www.omg.org/spec/BPMN/2.0/

Possible mentors

Isuru Wijesinghe (isuruwi AT wso2 DOT com)

Chathura Ekanayake (chathura AT wso2 DOT com)

Proposal 20: [IS] SCIM 2.0 Support to WSO2 Identity Server.

System for Cross-domain Identity Management (SCIM) is an open standard for automating the exchange of user identity information in cloud-based applications and services. It is designed to make managing user identities in cloud-based applications and services easier.WSO2 Identity Server is an open standards based Identity and Access Management system. Currently, Identity Server partially supports SCIM 1.1specifications. The objective of the project is to improve the existing SCIM implementation to support SCIM 2.0 specifications.

References