Understand this for me Software for Concept Analysis David Upton - PowerPoint PPT Presentation

“Understand this for me” Software for Concept Analysis David Upton (Stirling Reid Limited)

Operational analysis example

Words.

Language Games

Wittgenstein: Language Games Please paint the back door this afternoon….. Was that not what you wanted?

Ranges of meaning: “affordability”

Language games as group indicators “…We [SBS] even had a different language from the SAS. For scoff we said scran. We had a wet, they had a brew. We went for a yomp, they a tab. We went to the head, they to the *******…” (Duncan Falconer, ‘First Into Action’, 1998)

Operationalising this idea

Spies and cyphers

Noisy channel examples (1)  You receive ‘yhe’ instead of ‘the’:  Error explained by ‘y’ key being next to ‘t’ on standard keyboard.  Deduction: channel involved QWERTY keyboard.

Noisy channel examples (2)  You receive ‘ind’ instead of ‘and’  i = ..  a = .-  Error explained by one Morse code ‘digit’ being wrong.  Deduction: channel involved morse code

We need something to compare

Selecting a corpus?

Dictionaries Buildings? Large craters? Small craters? Roads?

Just a matter of counting! for file in target_list: filepath = self.training_folder+'\\'+file mydict = self.open_json(filepath, values) for item in mydict['data']: if item[0] in fuzzy_dict.keys(): fuzzy_dict[item[0]] += item[1]/mydict['size'] else: fuzzy_dict[item[0]] = int(item[1])/int(mydict['size' signature1 = {} signature1['data'] = fuzzy_dict.most_common(signature_length)

Ranges of meaning: “affordability” cost expenditure money media time reputation speed image late risk survival waste secure responsible bargain energy saving cheap

Summary

Summary • A system to measure the ‘meanings’ of words, as they are used by different writers • Avoid misunderstanding • Address requirements more precisely • Assess opposing views more clearly • Better document triage. • David@simulation.cc • (+44) 7957 597904