コミュニティ

Hello,
I would be very grateful if somebody could clarify any of the following questions..
1) Profiling/ Speed measurement in FLS:
1.1) Are there any profiling tools in FLS? When layout matching takes a long time it can say smth like "Element XX matching takes too long..Try to optimize search, i.e. restrict search area".
So I can assume that it has some time measurement mechanism, can it be somehow used in scripts?
1.2) Does FLS have global time/date functions and date manipulation routines? For manual matching profiling...(one workaround-is to use SQL TIME_STAMP/GetDate() functions)
2) Hypotheses generation/element properties:
2.1) is there a possibility to know the number of generated hypostheses so far of some element? The goal is to dynamically decide the number of possible hyppotheses
based on some calculations.
2.2) Do elements have access in scripts to its base properties like: element name, hypotheses count, null hyp coef, pre-search quality,post-search quality, chain quality?
2.3) Is it true, that current hypotheses generation mechanism is FLS is "optimized", that means it has some heuristics to minimize number of calculations? If so, is it still
equal to direct "full hypotheses generation" approach or not equal? Does anyone have a description of hypotheses generation algorithm?
3) FC post-processing questions
3.1) Is it possible to understand if current recognition hypothesis was a vocabulary or not?
3.2) Is it possible and how to check if the hypothesis passed a "value checking" procedure (the setting in the bottom of fileld's setting page that restricts value by max length,range of possible values or regex)
3.3) Is it possible (and how) to get other recognition hypotheses a) by words b) by symbols? Is it possible to get confidences a) of each word hypothesis b) of each symbol hypothesis?

Thanks in advance to all!