Dictionary Interviews with Mappers

TL’DR

 

A couple of mappers use the dictionary all the time. Most do not but when they have to approve the terms, they obviously do. When used, it is during tagging.

It seems we get most value out of the dictionary when terms are non-common and specific to compliancy or the customer industry.

A lot of claims are being made about dictionary terms being needed as ‘evidence’ but these claims were not substantiated. Mappers who do work with customers describe the process of agreeing with customers much more as collaborative where customers are offered choices between requirements/citations and then these are discussed.

What sources do they extract information from?

We get an idea by looking at which browser tabs are open when Mapping is being done.

 

The current project/document +

  • Research - a couple of tabs open

  • Dictionary

  • The source regulations, possibly in different versions

  • Previous document with mappings, in-depth reports

  • The Magic spreadsheet

I always have citations, common controls, and dictionary.

What steps are taken in the process of mapping?

 

Tagging

Sentences are tagged.

“what it helps us do is remove the noise, you can just see the comparison between the amount of text that's not tagged and then the amount of text that's tagged. One thing that this can help everyone do is more easily identified what exactly needs to be done."

Matching

The system (?auto - ?how) matches a sentence to a Common Control.

So this tries to do it for us. It's I don't know. This search doesn't tend to work for me.

I really never look at those results because I think because of what I mentioned, because it's not searching the right things.

 

A tool that helps with human verification of the match between Requirement/Citation and Common Control is the simple word match visualization:

 

 

When do you use the Dictionary?

During the Tagging process (3x/2)

Dictionary stuff would have come up. Let me think. When I was actually inputting tagging.

when I am in the matching phase, I don't really consult the dictionary

I don't truly use the dictionary a lot

When I was editing term relationships for matching control, I just go in here to make sure that I'm finding the right term, the right standard term. So that's what I usually do in in dictionary.

I just use it all the time, actually. Just because I feel like the term relationships aren't always they're they're not always conducive with what either me or the approver thinks would be the, you know, best match. And so And at 80 percent, you need a decent amount of connections.

When the tagger searches for words.

Eg to compare the words ‘patch panel’ and ‘network cabinet’ - will look up their definitions.

You know, you you might think it's one thing and then, in fact, it's the other thing when you see the exact definition and resources and what actually means.

Deeper definition matching happens because we are not industry experts, we are generalists.

Or when the system tells us there is no definition and we should add one to our dictionary.

This is doesn't exist in the dictionary. So I'm gonna create a new dictionary task

So 1 way I use the dictionary is when I'm creating terms, which isn't isn't a whole lot

(note about 10 times in 5-6 months)

Or, to review terms from mappers (approver role)

I I use it constantly

Using external sources

After going to research, and if nothing is there - go to Google

then is this a nonstandard term of existing term? And so here is where I need to go out and do some research for what this term means.

what I would do is go out and the world and search for both And then pull up a reliable dictionary.

And then I would try to find a way to merge these into a single definition that makes sense.

But ChatGPT seems to work well too:

I'll go into chat chat GPT and say, could you merge these into a dictionary definition, and and it does a very good job of that too.

What is important in the Dictionary?

most important part to me is just the definition is the, you know, central part of a dictionary just seeing the definitions and then seeing the sources.

Type:

Definitions type, whether it's a asset, or something other that is let me think. So it's important. In so far as if the mapper is trying to tag, for example, the word cable, and then the definition type is not see. Let's say if just that they're trying to just type the word server and the definition was not an asset that would give me reason to pause and make me wonder whether the definition was correct or not

It seemed hard to come up with real examples as to when type was useful

But I guess definition type could help us in so far as we can easily distinguish between terms and I don't have no no terms are coming to mind immediately.

clients have talked about how it's not helpful’

‘I've been working with people who are confused about it

‘it's nice to have a reference’

‘difficult to understand’

‘process is difficult to understand

 

Lexicographer has this process:

-make sure that everything's mechanically correct grammatically correct

  • make sure that that is the correct designator too.

 

I do love about the dictionary is we have terms that are defined nowhere else online’

Golden nugget

can't say for sure that clients don't don't already know what the words mean, but if if just in case they don't, we would like to be the the people we we we like to be the company that really offers them that clarity.

Sometimes some specific terms are so technical and any sort of information either it shows up very little in in online searches, or every instance of it is so technical that it just it's written for an audience that they assume already know what they mean by that.

How useful is the Dictionary?

definitions you're getting are just a general you know, the definition of description and the general definition of system, whereas if you tag this exact term right here, and or, you know and usually, you'd create the term if you create the term and give it a specific definition. If you give it a definition type record example, then what you're doing is making it.

 

<as evidence when customers ask/discuss>

that's why it's supposed to be actually useful because we could see we could weigh the evidence that the completed project will you know, that will be available and accessible in the completed project.

<note: ‘supposed’ and 'could' in the sentence above>

How do you use Relationships in the Dictionary?

For example, there was let's see. So, previously, there's a control about processing domains, and we were trying to decide whether and how it relates to server And so when I look at processing domain, just by looking at the definition, it's just saying distinct areas like a of computer processing within a server software application or system. And then I look at the immediate relations I didn't see anything that was directly related to okay. So it is directly related to server.

 

it's just good evidence if it's has plenty of relations

it's evidence that can be shown to a customer that our conclusions on his citation are in fact valid.

Note: interviewee could not validate this

So which customer did you show this to and that you have at reaction is, oh, that's how it works. Do you remember?

Not off the top of my head.

 

Other people don’t use it.

No.

(for general terms) this thing is huge.

for terms where it could be helpful, it's it's pretty bare just because it's like, the less relatable, I guess, the term is the fewer relationships it's gonna have.

They only look at this because the task asks that they add to the diagram.

go through this process and adding the definitions. But, really, that's the only time I would look at this.

The semantic relationships are very important, then that's also something that we cite to clients as 1 of the reasons that our control matches are so scientific.

 

But I didn't see that at the time

But normally, when we see something with s p relations as this, it's we need to send it back and we need to add at least 20 to it.

 

Re the need for evidence

There was one instance in four years where a customer disagreed

And the customer subject matter expert did not agree with the common control. And I forget forget exactly which common control it was

we have either online sources or we also have a per ISO source, which is based on a document that was mapped originally in 2009

More during customer projects

VMware, Verizon, AT and T, These big companies is subject matter experts as Boeing and stuff like that. It's a pretty There's a lot of dialogue and there's a lot of customer, how do I say, skepticism of our choices and disagreement with our choices.

 

"In those cases where we're dealing with particular particularly skeptical or demanding subject matter experts when things like all of the evidence available in the dictionary and even in the asset tables anywhere we could find in our resources in these recitations, everything here, it becomes you know, I mean, that's basically darsenal, and we have to take information from there and turn it into digestible information or just bullet points from which we can argue the validity of our choice."’

 

his was our original match, and they weren't satisfied with that. So this is AT and T notes, kind of explaining the purpose of the citation, what it's actually saying.

 

So then we would come up and say, okay. Well, here's 4 or 5 other controls that you have a choice of. You know? Do you like any of these? And then, basically, it would say yes and no. If they said no, then you say, okay. Well, you know, no. But this one's close. We need something a little bit more about this. And then either we would do it just on spreadsheet and go back for it there or, like, AT, we actually go through in a meeting and talk with them on Zoom and then, you know, then they'll go go go through line by line. They say, okay. Well, we like this this.

What is proper evidence?

But I I prefer to have sources, and I'm not talking Wikipedia. I'm talking actually proper sources.

a technical dictionary, like, a missed dictionary, or an isodictionary or even even certain configuration literature

 

But I will use those for Common? I'm sorry. Yeah. That's probably fair with, but, yeah, more common terms, you know, more commonly used terms, less complex terms, noncompound phrase, non compound phrases. I'll use those a lot. I'll I actually use those quite a lot. Webster, Oxford, Cambridge. I sometimes use Word, Nick, And that's just for the more common terms. But for the more technical terms, I definitely do try to defer to either more technical sources, potentially multiple sources, or I have to sort of craft a definition from from research.

Re the Assets

I feel like all of these element tables don't know how much people really even search here, but, you know, mappers enter all of this information.

And so I feel like this is it's interesting information. And, you know, it took us time to enter this, but I don't know if anybody will ever look at this again.

 

Re audits

Audits are often mentioned, but we lack evidence ourselves as to what specifically an auditor needs and wants

I've I've never actually been engaged by an auditor before.

Why we create a spreadsheet

Mappers created a spreadsheet. So far, every single mapper we spoke to was using this.

And because sometimes met citations are really similar, but you can't really find them using the messaging portal or using the mapping project. But over here, I could just, you know, search for something that comes into my head, and I could just do a control f search And then I just have everything every matching task that I did for the whole project right here. So I could just do a control of the search, and then I could find the suggestion that I want relatively quickly. And this also checks all of the I keep all of the mapers' choices here. I