You can write an essay on all of your questions, but let me try to be brief:
1 - Difference between Citrix automation and Computer vision Recorder?
You use Citrix automation to retrieve selectors from elements in citrix application or virtual desktops. This enables you to use regular UiAutomation activities. If you cannot retrieve reliable selectors you might want to use Computer Vision from UiPath instead to identify elements properly. See referenced documentation articles.
2 - Run specific queue item
You can give QueueItems a reference and retrieve them by that reference with Get Transaction Item activity or Get Queue Items activity. The latter has even more options to filter the query. For very complex cases, you can try to use an Orchestrator HTTP request. Search the forum for example solutions!
3 - What is relative scraping in UiPath?
The documentation says it all:
Relative Scraping is a technique that enables you to retrieve text from UI elements by using OCR technology. In situations where selectors cannot be found, the target UI objects are identified by using image recognition activities to look for adjacent labels or other elements.
4 - In OCR which is best one to extract the data
In my personal experience the trophy goes to ABBYY OCR (I think they are a UiPath partner too)
5 - What is serialization and De-Serialization? what is the use of both?
I’m not an expert on this, maybe someone else can elaborate. All I know is some objects are serializable and some objects you can deserialize. Example: Deserializing a Json string into a JsonObject.
Hope this helps!