Natural Language Processing (A.I.) in Software Testing.

Software testing and A.I. - TECHSAGE europe has been working on a research project to see where A.I. can be of help for software testing.  

The Projects started as a hobby project during Corona period picked up momentum when the results were promising. (Below is a demo video to show the very simple A.I NLP based Test assistant at play.)

Building a custom NLP for Testing. But what is NLP, many will ask.

Natural Language Processing (NLP) is a field of artificial intelligence (AI) that focuses on the interaction between computers and human language. It's a branch of AI that enables computers to understand, interpret, and generate human language in a way that is both meaningful and useful.

NLP involves the development of algorithms, models, and techniques that allow computers to process and analyze text and speech data. These algorithms enable computers to perform various tasks that can be relevant for software testing, such as:

1. Text Analysis: 

NLP can analyze and understand the structure of written text. It can identify sentences, paragraphs, and individual words, and it can also determine the grammatical structure of sentences.

2. Language Understanding: 

NLP helps computers understand the meaning of words, phrases, and sentences in different contexts. This includes recognizing synonyms, antonyms, and the relationships between words.

3. Speech Recognition: 

NLP enables computers to convert spoken language into written text. This is used in voice assistants like Siri and Alexa, as well as in transcription services.

4. Information Retrieval: 

NLP helps search engines understand user queries and retrieve relevant information from vast databases of text.

4. Calculations and complex tasks: 

NLP helps execute some important calculations and tasks which might be error prone or their could be chances of making mistakes when done repetitively. So relying on an algorithm to do those for the user is a good way to reduce user errors.

8. Reporting Generation and data comparisons: 

NLP can be used to generate test reports with no human intervention. Learning from the historical runs and comparisons can also be done and placed in the end to end cycle to compliment the ease of testing.

Below is a Demo video of the recently added Natural language processing (NLP) Test Assistant which makes lives easier for Test Engineers, performance test engineers, Managers, Test managers and actually also for all other team members.
 

This Test assistant has made the task of Test execution very easy and simple for any Agile team member. What they only need to do is say a voice command and name the test which they want to run.
The Assistant takes care of the following actions:
1. Listens and recognizes the voice command.
2. Identifies the test.
3. Starts authentication and Security checks.
4. Setup test environment.
5. Locates Resources like scripts, test data etc.
6. loads the correct test script.
7. Adds dependencies - Activates real time dynamic input messages creator.
8. Loads the Load models available for the test.
9. Connectivity checks...
10. Starts the test execution.
11. After the test execution is complete. Creates the Test Reports.
12. Emails the test reports to the person who started the tests.
13. Cleans up and again ready for the next command.

It might be useful when for example the test engineer is on holiday or sick, or has some other more important tasks to finish. Anyone from the agile team can take up this task of test executions.

A non technical team member can also run the test by just knowing the test name and the command. 

There is no need to create the test environment, dependencies, configurations. It is taken care by the platform.
To run a test all you need is a browser on any device. It could be a mobile/laptop/tablet or a smart watch. Refer the video below for a short demonstration:
for more information mail to info@techsage.eu