Artificial Intelligence in Enterprise Applications with TensorFlow and Joget DX

By -
 
Joget DX is an open source no-code/low-code application platform that introduces innovative features such as automatic support for progressive web apps (PWA), integrated application performance management (APM) and built-in artificial intelligence (AI) support. 

In the previous article A Quick Introduction to Artificial Intelligence, Machine Learning, Deep Learning and TensorFlow, we introduced the concepts behind these terms, and how things generally work. In this article, we’ll have a look at how to incorporate TensorFlow for AI use cases on the Joget platform.


AI and Decision Support Features in Joget DX

To simplify process automation, Joget DX supports Decision plugins that can be mapped to process routes for decision making. Several implementations are bundled, including a no-code Rules Engine and a no-code TensorFlow plugin to execute pre-trained models.



The AI focus in Joget DX is to simplify the integration of pre-trained AI models into end user applications. As rationalized in the previous article, the training of AI models are best left to machine learning experts so once a trained model is available, the goal is to make it as accessible as possible to app designers.

With the bundled TensorFlow AI plugin, you essentially:
  1. Upload a pre-trained TensorFlow model exported in protobuf (.pb) format
  2. Configure the inputs and outputs
  3. Configure optional post processing
The following sections showcases how a sample app on Joget DX incorporates some well known models for several common AI use cases:
  1. Image Classification
  2. Object Detection
  3. Face Recognition
  4. Audio Classification
  5. Text Sentiment Analysis

Sample No-Code AI Apps on Joget DX

Image Classification

Inception v3 is a widely-used image recognition model that has been shown to attain greater than 78.1% accuracy on the ImageNet dataset. ImageNet is a dataset containing for image classification containing than 14 million labeled images. In the sample app, a simple decision process is designed to demonstrate the use of the TensorFlow AI Decision plugin as shown below:


 The TensorFlow plugin is then configured by:
  1. Uploading the pre-trained model in protobuf (.pb) format
  2. Configuring the image with appropriate dimensions as the input
  3. Configuring where to store the output result
  4. Configuring additional post processing to match the output result to a specific label


 Download the model from http://download.tensorflow.org/models/inception_v3_2016_08_28.tar.gz. More information on the Inception model can be found at https://towardsdatascience.com/a-simple-guide-to-the-versions-of-the-inception-network-7fc52b863202

Object Detection

SSD MobileNet v1 is a model for object detection trained using the COCO dataset. COCO is a large-scale dataset for object detection that contains 1.5 million object instances. The model can detect and identify multiple objects in a single image, as shown in the image below.


 The TensorFlow plugin is configured by:
  1. Uploading the pre-trained model in protobuf (.pb) format
  2. Configuring the image with appropriate dimensions as the input
  3. Configuring where to store the output results
  4. Configuring additional post processing to match the output result to specific labels


 Download the model from http://download.tensorflow.org/models/object_detection/ssd_mobilenet_v1_coco_2018_01_28.tar.gz. More information on object detection can be found at https://www.oreilly.com/ideas/object-detection-with-tensorflow


Face Recognition
FaceNet is face recognition model introduced by Google that converts a face image into 128 measurements called an “embedding”. One embedding is then compared to another in order determine whether or not a face matches. This sample is based on the sample at https://github.com/davidsandberg/facenet.

The TensorFlow plugin is configured by:
  1. Uploading the pre-trained model in protobuf (.pb) format
  2. Configuring a couple of images with appropriate dimensions as the input
  3. Configuring where to store the output results
  4. Configuring additional post processing to compare the “distance” between the embeddings to determine a match


 Download the model from https://github.com/davidsandberg/facenet/wiki/Validate-on-lfw. More information on face recognition can be found at https://medium.com/@ageitgey/machine-learning-is-fun-part-4-modern-face-recognition-with-deep-learning-c3cffc121d78

Audio Classification

 ResNet stands for residual network, and as the name implies, it utilizes residual learning to preserve good results in neural network layers. ResNet demonstrates good results for image recognition, but also shows promise for audio classification e.g. determining the musical genre for an audio sample. This sample is based on https://github.com/chen0040/java-tensorflow-samples/tree/master/audio-classifier

The TensorFlow plugin is configured by:
  1. Uploading the pre-trained model in protobuf (.pb) format
  2. Configuring the audio sample with appropriate dimensions as the input
  3. Configuring where to store the output result
  4. Configuring additional post processing to match the output result to a specific label


 Download the model from https://github.com/chen0040/java-tensorflow-samples/tree/master/audio-classifier/src/main/resources/tf_models. More information on audio classification can be found at https://medium.com/@CVxTz/audio-classification-a-convolutional-neural-network-approach-b0a4fce8f6c

Text Sentiment Analysis

CNN for Sentence Classification is a model for text classification, for example to determine whether customer feedback is positive or negative. This sample is based on the sample at https://github.com/ivancruzbht/tf_android

The TensorFlow plugin is then configured by:
  1. Uploading the pre-trained model in protobuf (.pb) format
  2. Configuring the text with related dictionary as the input
  3. Configuring where to store the output result


 Download the model from https://github.com/ivancruzbht/tf_android/tree/master/app/src/main/assets. More information on text classification can be found at https://medium.com/jatana/report-on-text-classification-using-cnn-rnn-han-f0e887214d5f

Getting Started on AI with TensorFlow and Joget


Joget DX will be released later in 2019, and the sample Joget app containing these AI use cases will be available then. In the meantime, TensorFlow can be incorporated into Joget apps using a custom plugin as described in an earlier article. The current stable release of the Joget platform, Joget Workflow v6, is available now for download or on the cloud.

Comments

  1. doradolearningSeptember 22, 2019 at 3:02 PM

    This is the first time I came to know about Joget DX and its use in PWA Apps . I will keep following your blogs for more such contents.

    ReplyDelete

Post a Comment

Popular Posts

Security in Low-Code Applications: Cybersecurity Features and Best Practices for the Joget Open Source Platform

By -
Image
Introduction to Security Benefits and Risks in Low-Code Platforms As digitalization becomes increasingly prevalent, low-code platforms are rapidly gaining adoption across industries. Employing visual techniques to programming, they enable the development and delivery of applications in a fraction of the time required by traditional coding. According to Forrester’s Now Tech: General-Purpose Low-Code Development Platforms, Q1 2021 , low-code platforms are “vital to digital business” and “proven in even the most mission-critical use cases”. However, the proliferation of apps and digital solutions bring with it increased cybersecurity risks like phishing attempts and ransomware attacks. As the SolarWinds hack in 2020 and Microsoft Exchange hack in 2021 demonstrated, the cost of security lapses can be monumental. Security considerations are therefore paramount, so how would they apply to low-code platforms? Source:  Don’t Ignore Security In Low-Code Development, Forrester Research Inc...
Read more

Transforming Customer Onboarding: How Allied Benefit Systems Achieved Reporting Accuracy and Streamlined Business Processes with No-code/Low-code.

By -
Image
To optimize business operations and ensure accurate reporting, Allied Benefit Systems empowered business users and delivered a Customer Onboarding Platform built on no-code/low-code within three months. Effective onboarding processes play a crucial role in creating a positive impression on new customers. Yet, many businesses still need to rely on manual methods that can be time-consuming, error-prone, and inefficient. Allied Benefit Systems (Allied) , a US-based healthcare solutions company that prioritizes people, understands the importance of streamlined onboarding. As the largest independent third-party administrator in the US, Allied designs tailored self-insurance benefits solutions that integrate innovative medical management and cost-control strategies.   The company collaborates with clients and benefits consultants to institute plans that support the best options for employees and their families. However, Allied's existing onboarding process was managed through spreadsheet...
Read more

No-Code, Low-Code and Pro-Code: Why Flexibility is Essential for Digital Transformation

By -
Image
Introducing Business Reporting and API Management Addons for Joget DX The COVID-19 coronavirus pandemic will be viewed in history as a significant point in human history. In the post-pandemic world, digital transformation can no longer be delayed and organizations need to adapt quickly, or perish, as the old ways of working no longer apply in the new normal. Based on Forrester’s Q1 2020 Digital Process Automation Survey , over 70% Of organizations still have paper-based process dependencies, as shown in the figure below. With a modern open source no-code/low-code platform like Joget DX , organizations can rapidly build, deploy and adapt applications to cater to ever changing needs. In this article, we will talk about the need for flexibility in application platforms to meet digitalization needs, and how the extensibility of Joget DX enabled the introduction of new add-ons for business reporting and API management . No-Code, Low-Code and Pro-Code: Flexibility and Extensibility There mi...
Read more

DISINI™ Vehicle Inspection & Insurance App

By -
Image
The DISINI™ Vehicle Inspection & Insurance App is an application that can be used standalone or in a composable application environment on the Joget DX low-code platform. Enterprises will be able to use the application to manage their fleet of vehicles, track insurance policies expiry and conduct regular vehicle inspections. Summary of the App features are: Track your vehicles' insurance policies and set email reminders of the renewals. Assign inspectors to conduct scheduled vehicle inspections. Supervisors will be able to set reminder notifications for the inspectors of upcoming vehicle inspections. Inspectors will be able to check their calendars for scheduled inspections that have been assigned to them. Create and manage inspection checklists templates for the many vehicle types. Inspectors will be able to update and check their Dashboard for inspection tasks. Supervisors will be able to keep track of all assigned inspection tasks and follow up accordingly. The following ar...
Read more

Superapps in the Enterprise: Up, Up and Away!

By -
Image
 “We think these applications are so compelling that we don’t just call them apps, we call them super apps… Super apps are applications that once you start using them, you’ll wonder just how you lived without them” - BlackBerry founder Mike Lazaridis. What is a Superapp?  In the early days of the smartphone revolution, BlackBerry founder Mike Lazaridis first coined the term “super app” (or “superapp”) in a MWC 2010 keynote speech . He defined them as “a closed ecosystem of many apps that people would use every day because they offer such a seamless, integrated, contextualized and efficient experience.” According to Gartner®, “a superapp is like a Swiss army knife with a range of component tools (miniapps) that the user can use and remove as needed”.  We consider superapps to be a single platform where miniapps are integrated, thus providing a seamless experience to its users. Here are some statistics and predictions from 2022 Gartner Top Strategic Technology Trends for 2...
Read more