A note on user documentation

Usroc is topologically equivalent, but not simpler. Usrdoc to a nine-year-old abacus, spherical, of course.

Keywords: controlled language, faceted search, no hypergraph , software development, technical communication

Is an introduction really necessary?

(*) Dear Santa, all I want for Christmas is usrdoc. This will be not effective.

Contrary to other deliverables end-user documentation for a software system is frequently not build at all. On the other hand, value could be created . Given it exists, user documentation is far from optimal (1) Remember: “The system must be user-friendly!” (2) One of many existing reasons for the lack of  loser documentation is that, at a second glance, implementation becomes quite demanding. Various aspects are to consider (**). True, software development projects differ (***). Overall however usrdoc has to be pictured as simple as possible, but not simpler. (****).

Footnotes:

• * Note: All URLs were active on date of most recent retrieval. Average time: depends.
• ** e.g.. convenient UX increases acceptance
• *** e.g. by the potential end-user
• **** As coined by the famous American scientist Einstein.

References

How can it be defined?

End-user documentation for a BI system is called usrdoc here.

References/Top

What does this mean?

Usrdoc is question answering, as using the system questions, which deserve an answer, occasionally will arise. It is web help (the method, but not the tool with a similar name).

What is itnot?

• Online help. Mostly RoboHelp is recommended for the creation.
• ID does not belong to usrdoc. Though, many issues are of IT-technical nature.
• E-learning. Learning should happen nevertheless. Thus, diverse approaches must be offered. (3)
• IAEFRTM. From the proven DocBook to the more theoretical DITA, there are many options to create conventional manuals.(4)

What are the consequences?

References/Last/Usrdoc/Top

Which criteria apply to usrdoc?

1. clarity
2. usefulness

Contrary to usrdociotism,  this is easy to remember. Just think of “Cyber Unit“.

References/Last/Usrdoc/Top

How is the language constructed? (5)

The headwords from the General service list and the Academic word list (*) are collected, memorizing their classes (6). (**) Numbers are added. There are building rules to obtain new words (***)). For sufficiently frequent words, grammar is made use of. Let YAESP be, roughly defined, the name of the resulting controlled natural language. YAESP is supposed to be part of the natural language English, except for illogical or local conventions . Please consider that notwithstanding a transition from Standard English to YAESP is a shallow translation. Based on the current, stable procedure (****), NLP is interesting, but of minor help (7)(8), as, for instance, resolving the ambiguity of single words (*****) and building sentences (******)  are topics in research. Though, there is a cavalry near to you. Back to square 1.  Probably. even a commercial corpus has undergone manual data “cleansing”, for the errors (9), according to the obscure personal taste of an editor. Did you know that Robust statistics should be applied here?

WARNING: The above is not intended to be certified advice. For this purpose consult a linguist near you!

Footnotes:

• * Surprisingly, obfuscate is not included, but “prospective” is.
• ** As a list of words is not sufficient for automatic processing, see e.g. how to order the free DCPSE treebank.
• *** e.g., to prefix an adjective by un
• **** e.g. peer review
• ***** John views the girl with a telescope.If you buy the enterprise-version of a certain software , translation is all easy. Next release will support Korean.
• ****** Will the real sentence please stand up? How do you know that a group of words is really a sentence? So much for punctuation.

References/Last/Usrdoc/YAESP/Top

Is domain knowledge incorporated?

Yes, but not automatically. Expertise may be present within a considerable range. Sometimes it is contained in flat files, sometimes in databases, in an arbitrary information repository or hidden in application software. Thus, the way of harvesting or mining  depends. Occasionally, there will be conflicting results. To feed intermediate  is not again scheduled to be an company-wide activity. Google is the friend of everybody (apart from Mr. Wolfram) (11). Probably Wolfram omega, due to the GIGO principle, will still ask to specify words instead. This kind of search is done, as the perpetual three steps of prearrangement for an open search:

1. crawling a small (*) part of the web
2. Indexing “words” for better performance of a potential later retrieval
3. applying an obscure, ever-changing algorithm (**) for determining the sequence of the presentation of all the impressing 5 million results

leads to a disappointing  (***) outcome.

Footnotes:

• * e.g. not the Deep web, of course
• ** SEO is all about guessing this, yawn
• *** You can disappoint some people all of the time. You can disappoint all of the people sometimes. You can not disappoint all of the people all of the time. At least, this will not become a success story, then.

References/Last/Usrdoc/YAESP/Top

Are arbitrary questions answered?

Never (*) all conceivable, no valuable if no specification by formal methods exist. (**) Should I ask Oscar instead? Back to the shoulders of other giants. (12)(13)   A certain generalization of a graph is popular to store corresponding concepts, reasoning and entailment afterwards. Although, often (business processes (14) to representation theory (15)) issues are handled differently, this approach would guarantee that there are many tools at hand.

Footnotes:

• * no philosopher helps
• ** John has 2 cousins. How many cousins does John have? 42, maybe.

References/Last/Usrdoc/YAESP/Top

What are the components?

Last/Usrdoc/YAESP/Top