Homework Responses Week 6

Good morning everyone. I hope you are all having a wonderful week. I cannot believe we are almost done with week 6.

I based my experimentation process from past research conducted by Stefan Cremer and Claudia Loebbecke, published instruments.  Within their project, they are trying to show the need to integrate artificial intelligence within imagery analysis. The Azure Cognitive Services from Microsoft are able to contain applications encompassing “decision-making, speech, language, vision and search” (Cremer, 2019). Similar to humans, this type of AI has the ability to process facial recognition, emotion, image classification, etc. With the services from Microsoft, they are able to harness the power of all those skills, and produce a product with all of the data and descriptions. The control here are the images that will be uploaded to be analyzed. They are generic thumbnails taken from certain resources to provide good experimentation data. With my argument, I am reaching that this tool should be used across the intelligence community, so if I were conducting this experiment, I would change the pictures to be that of SARs, sensor generated, etc. That would be a modification in terms of tools. I would also implement analysts with similar qualifications and experience to see what they pull from the pictures, and do a comparison.

The data being collected is contains both quantitative and qualitative information. The time it took to process, what it did process, and how it categorizes the picture can be pulled from the software application being uses. In terms of validity, the comparison of the human generated analysis and the artificial intelligence generated analysis, it would center around criterion validity. In order to prove the content validity, there would have to be some research conducted and published within the research paper about the applications from Microsoft. For research reliability, since I am basing this experiment off of experiments done before, I can see what consistent results work, and what issues arise in the process. Test-retest is another instrument that would be useful, to ensure that each decomposition of the pictures is true, and that the data does not change.

Works Cited

Cremer, S., & Loebbecke, C. (2019). Artificial intelligence imagery analysis fostering big data analytics. Future Internet, 11(8), 178. doi:http://dx.doi.org.ezproxy1.apus.edu/10.3390/fi11080178