[Slightly OT] Jeremy speaking at CodeMesh, London 8/9th November 2018

[Jeremy Ruston]

Joe Armstrong has kindly invited me to speak with him on "Intertwingling the Tiddlywiki with Erlang” at the CodeMesh conference in London on November  8/9th 2018. For those that may not know, Joe is the eminent creator of Erlang, a computer language that has become dominant in real time telecoms systems over the last 30 years. Joe has been interested in TiddlyWiki for a few years, and recently we’ve been collaborating on ideas that emerge from mixing the concepts of Erlang with those of TiddlyWiki.

I’m afraid it’s a commercial conference with a hefty price tag, but may be interesting for anyone with a supportive employer who funds conference attendance. If it’s not recorded I think Joe and I can undertake to repeat the talk for posterity at some point.

https://codesync.global/conferences/code-mesh-2018/

I don’t normally seek out opportunities for public speaking, but I’m quite excited about this one. It’s been fascinating to see how Joe thinks about TiddlyWiki, and encouraging to have some mainstream interest.

Best wishes

Jeremy.

[@TiddlyTweeter]

Jeremy Ruston wrote ...

<snip> ... It’s been fascinating to see how Joe thinks about TiddlyWiki, and encouraging to have some mainstream interest.

IMO this is good to get TW more notice. TW's Flexible Utility is, I think, quite unusual.  Interested to learn about what happens, how its presented/received.

Best good luck,
Josiah

[Alex Hough]

Jeremy,

I have an idea! 

Could we make a separate artefact with you and Joe. I am kick starting a not-for-profit project, it has a Code Club as part of it.

Thanks to Joe's Tweet mentioning Sonic Pi and TiddlyWiki as great starting points for learning to code, we are planning an event using Sonic Pi

maybe we could take this converstaion off line before reporting back to the TiddlyWiki community... 

 

best wishes

Alex

--
You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+unsubscribe@googlegroups.com.
To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at https://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/A79BC829-DC1B-481F-8F9F-77B2B8FC32F1%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

[barro...@gmail.com]

Definitely, take TW wherever it can be taken to.

[PMario]

Hi,

Here's the video: https://www.youtube.com/watch?v=Uv1UfLPK7_Q&index=9&list=PLvL2NEhYV4ZtWFBNOrApXaIoCTtj-yk7Y

have fun!

mario

[@TiddlyTweeter]

I was surprised. Its more interesting than I thought it would be! :-)

Jermolene and Joe Armstrong are developing some ideas that are both interesting and kinda look doable.

Various concepts were discussed. Here are couple of several ...

I found interesting the general idea around ENTROPY REDUCTION. Meaning there is too much unneeded repetition in lots of knowledge stores. The point was something like the SEMANTIC UNIT in TW better corresponds to the section of text (paragraph-ish) that is such a unit than is the case in word-processors--where documents subsume semantic fragments into "coerced wholes" that need "difficult decomposition" to "un-discombobulate" (my three phrases). This could have useful implications for writing more efficiently in TW ... for example ...

1. Why on earth would I want to write a Tiddler about a Toothbrush Inicident I had if Mark S. has written about his already? ...
2. In writing my Deep Thoughts I'd like to know IF they have been written already. I'm of the mind to Save Energy rather than repeat what has been done before.
   

Another thing was what I now think of as non-centralised-TAG-HARMONISATION. They called it "tag inference". I think their point was if you combine or co-ordinate wiki from different authors they likely used different tags to organise Tiddlers. The idea was some kind of  algorithem to help co-ordinate them (Bayesian statistics seemed most successful in their early tests, which is itself interesting, though perhaps still not accurate enough.)

One thought would be that every tag could be peer-shared to form a THEASAURUS OF TAGOLOGY (my rubric) that maybe draw on existing online resources (my thought) like VisualTheasaurus to offer options live as you create tags...

Best wishes

Josiah

[@TiddlyTweeter]

One of the things that interests me a lot that the talk raised a bit--and which no one seems to know how to answer is ... :-)

- WHAT exactly is an SU (Semantic Unit) in TW writing (or computing writing In General, for that matter)?

There is a kind of rule of thumb "its maybe a paragraph"? But, of course that won't quite work for the one-sentence brevity of a Nietzsche.

Its obviously highly context dependent. And I doubt much of that context lives on the computer itself.

The idea in TW towards writing "the shortest semantic whole possible" (the word "fragment" here that is thrown around has muddied waters; they are not fragments so much as whole-parts-of-wholes) allows for later re-combinations to form more complex semantics.

However, I think its bit of an, ultimately, moot and mute point, in the sense that human meaning is often an interaction with technologies of expression themselves (though no where ever fully defined by them). So its an area of intuited understanding, not formal logic? On the other hand, who's offering the horse which water?

Josiah

On Monday, 10 December 2018 12:49:14 UTC+1, PMario wrote:

[TonyM]

Josiah,

Riffing on from what you said, here is a personal reflection:

A life in IT has taught me many things, once we become more expert at something, some of the basics become internalised, they take on an intuitive understanding rather than needing the application of intellectual rules, ie their cognitive load is reduced.  The use of system 1 not System 2 (see https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow) one of my favourite books this decade.

What I find with TiddlyWiki, it is somewhat unstructured and thus very flexible, its features allows one to make a mess as much as it allows one to exercise the skills and knowledge one has acquired. I seem to have no difficulty naming tiddlers and never seem to have problems renaming them because I rarely if ever have any dependencies on tiddler names, or where I do there is never a reason to rename them, nor do I seem to have any difficulties intuitively knowing exactly when to divide the content in anyway, that is, I seem to have no difficulty in understanding a  SU (Semantic Unit) in TW writing (or computing writing In General, for that matter)? The problem is this is already in my system 1 and  am often finding myself trying to reverse engineer this knowledge, so I can guide others towards using the same methods and rules. Sometimes I want to understand what underlies my intuition, to build a conceptual model, and sometimes extend its power.

I would speculate however my grasp of Semantic Units is based on the lessons of the following disciplines

• Structured software design and programming
• Object oriented design
• Analysis and Synthesis
• Database design and "normalisation"
• Alternate database models (Structured, Network, Relational etc...)
• Modularity and blackbox design principals

TiddlyWiki allows the democratisation of knowledge and application of algorithms commonly found in the above. But there are few rules.

I think we may need to obtain or construct a new discipline, that draws on the above disciplines (and others) selectively, such that we can pass to those seeking to apply knowledge and algorithms on top of our   "non-linear personal web notebook" or our "non-linear platform".

My use of TiddlyWiki continues to evolve rapidly, but I believe this is in part due to my understanding of the underlying concepts and patterns acquired in a life as a Information Technology professional. The question is how can we maximise what others, without such experience can do? or the exchange of such concepts between those with the expertise to others with similar needs.

Regards

Tony

[@TiddlyTweeter]

Ciao Tony

Fascinating comments (for me at least), thanks!

Your bullet points are interesting for all being (what I would call) "technical", apart from "analysis & synthesis". Its interesting how the word "semantic" has expanded. Though I was totally non-plused when I first encountered the term "semantic markup"... WTF is that?

In time I kinda got it. It leads to better HTML docs among other things.

In my post I was kinda riffing more about chunking of "end-text" content, not mechanism. Of course, any chunking is mediated by the medium/mechanism you are using. But the medium is not (I'm anti-McLuan) the message exactly. The human user always (at the moment) retains a Vorple Sword (thank god).

To try illustrate what I meant about semantics being "context dependent" ... say I made a TW of 4,000 Tiddlers quoting Perry Mason TV series ... that would be for the purpose of illuminating "how common tropes of TV are written."

Does my computer comprehend that or see it as semantically valid?

This bringing of "meaning for" to the TW tech seems to be a central issue where its flex gets very interesting--IF you understand it.

Something like that

Best wishes

Josiah

[Joe Armstrong]

Thinking out loud here ...

I've been thinking more about tags. One problem is that tags are rather vague and are written in different human languages.

One way out of this might be to adopt the wikidata word definitions. For example, I am, unambiguously

    https://www.wikidata.org/wiki/Q1691321

There are actually several Joe Armstrong's (for example, https://www.wikidata.org/wiki/Q712592)

These Q numbers uniquely define subjects and objects. Verbs (or predicates) are given by P numbers

so https://www.wikidata.org/wiki/Property:P178 means "the organisation or person who developed the item.

in RDF speak the triple

    {https://www.wikidata.org/wiki/Q1144644, 

      https://www.wikidata.org/wiki/Property:P178, 

      https://www.wikidata.org/wiki/Q17031730}

(BTW I recommend clicking on these links and playing around - there's lots of interesting

data in RDF tuples and the above links are a good place to start looking)

Means "TiddlyWiki developer Jeremy Rushton"

These triples encode facts in a hopefully reasonably clear manner.

So now the N$ question - can we automatically analyse a tiddler and turn it into a set

of RDF tuples. If we could then we could add these to the huge databases of RDF tuples

and possible find stuff in a clever way.

The filter notation in the tiddlywiki reminds me very much of prolog, and I guess with a but of

work SPARQL queries might be possible (SPARQL is an RDF query language)

Cheers

/Joe

[Alex Hough]

Hi All,

I've sent an email to Sonix [1] requestion time to transcribe the talk.

I also get free minutes if other people sign up using the link [2]

My grand idea that the audio could be cut up into semantic units and put into TW

I did a test [3] with a YouTube of Mark Fisher [4]

I wondered if the Sonix audio editor could be embedded in a Tiddler...

Once the text is in the editor you can use it a bit like a sampler. I[5]. Use the browser's built in search to navigate to word and phrases -- and hey -- why not feed it into SonicPi for your sonic art project?

Alex

[1] https://sonix.ai/

[2] https://sonix.ai/invite/kmnagvw

[3] https://sonix.ai/r/xpe6rwtXX8sm1pbMUP4okUe1

[4] https://www.youtube.com/watch?v=AhOcwhwumP4

[5] https://soundcloud.com/startdivergence/r26-0003-1wav

--

You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.

To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.

To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at https://groups.google.com/group/tiddlywiki.

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/15b60976-2b34-4947-abdc-b24cedd36584%40googlegroups.com.

[TonyM]

Josiah,

The so mominated technical items, are sourced from technical diciplins, however the relavant learnings to dealing with data, information and knowledge are not dissimilar to understanding analysis and synthesis. They include such concepts that are so importiant they should be understood publicaly.

One example is "normalisation" from database design and management. The pithy statement that comes from that is the details in a given record should be related to the key, the whole key, and nothing but the key.

Perhaps we could say that every field including the text field in a tiddller should be related to the tiddler title, the whole title and nothing but the title.

Understanding this could fix a lot of speadsheets out there or make obviouse, common logical errors.

Regards
Tony

[@TiddlyTweeter]

 Joe

Brilliant stuff! Most interesting.

One thought I had was you maybe inclined to the biggest picture, but at specific TW level harmonisation of tags it may be a more delimited issue involving less stress? :-)

The RDF triangulation looks good. But how far would I get? ...

human - male - born 19xx - London - Joe Armstrong

With a cross-cut I see you are

(not yet a transvestite :-) And you went to my university, so colleague. But were you in my excellent school (LSE)?

Back to TW ...

What users of TW do with tags in their wiki is not the same as a networked service would be. Why? because most of it happens in PRIVATE. And they use tags  for a several different reasons. For instance, some tags are used to aid generative ORGANISATION, rather than for semantic marking. And Special RESERVED tags are used to invoke system functions.

In other words Tags in TW serve several functions that can be orthogonal to each other.

TBH I need better grasp what aspect of tagging in TW you want to address. Something like that.

Just early comments

Josiah

[@TiddlyTweeter]

Ciao Tony

Thanks for this ...

TonyM wrote:

... Perhaps we could say that every field including the text field in a tiddller should be related to the tiddler title, the whole title and nothing but the title ...

Its interesting both the style of tagging and its various emoluments (profits). But the "title" is already a kind of default tag.

TW is interesting because its tags serve several functions (semantic, organizational, systemic) seamlessly.

But, at the same time, any TW tag is a "label applied" to a tiddler -- a distance between the tiddler and its manifest content.

FYI I'm a big fan of Twiitter where #hashtags are always inline. No separation of content from organization. Its a neat approach on content cognisance. Twitter is maybe extreme in its #hashtaggery but its effective in terms of finding stuff well enough. But, of course, Twitter usage of #hashtags is purely about flagging content, whilst in TW tags do several jobs.

Just comments

Josiah

[Joe Armstrong]

On Wed, Dec 12, 2018 at 5:10 PM @TiddlyTweeter <Tiddly...@assays.tv> wrote:

Ciao Tony

Thanks for this ...

TonyM wrote:

... Perhaps we could say that every field including the text field in a tiddller should be related to the tiddler title, the whole title and nothing but the title ...

Its interesting both the style of tagging and its various emoluments (profits). But the "title" is already a kind of default tag.

TW is interesting because its tags serve several functions (semantic, organizational, systemic) seamlessly.

Yes - I discovered this.I analysed a few TW's in detail.

What I did was to use Baysian inference to "learn" the relationship between the words in the text and the supplied tags - so for each word in the text I caculate the probability that the tiddler has tag <T> (forall known tags <T>) - then in a second pass I tested the model and predicted the tags from the text. This way I could correctly predict about 80% of the tags from the text alone. The problem was that, to me, many of the tags were meaningless and were used internally to organise the TW.

In a second experiment I totally ignored the assigned tags, and predicted the tags from

a TF*IDF analysis of the text. This made tags that made much more sence to me, but the

predicted tags often missed the supplied tags.

In my opinion the TF*IDF were better than the assigned tags since they had nothing

to do with the organisation, but more to do with the actual words in the text.

 

But, at the same time, any TW tag is a "label applied" to a tiddler -- a distance between the tiddler and its manifest content.

FYI I'm a big fan of Twiitter where #hashtags are always inline. No separation of content from organization. Its a neat approach on content cognisance. Twitter is maybe extreme in its #hashtaggery but its effective in terms of finding stuff well enough. But, of course, Twitter usage of #hashtags is purely about flagging content, whilst in TW tags do several jobs.

YES :-)  -- Given my earlier observations, perhapse we could distinguish two types of

tags. The #inlineHashTags could have something to do with the content of the containing paragraph. The tiddler tags could mean "tags used to internally organise the TW itself"

Cheers

/Joe

 

Just comments

Josiah

--
You received this message because you are subscribed to a topic in the Google Groups "TiddlyWiki" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/tiddlywiki/FXGEIa85Bks/unsubscribe.
To unsubscribe from this group and all its topics, send an email to tiddlywiki+...@googlegroups.com.

To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at https://groups.google.com/group/tiddlywiki.

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/9b2ed207-661d-45bb-92a2-cf049f5a2138%40googlegroups.com.

[@TiddlyTweeter]

Ciao Joe

A few comments ...

Joe Armstrong wrote:

What I did was to use Baysian inference

Top notch. (Side note: Bayesian stats are great, based on an ontology that is quite different than other approaches. IMO Bayesian approaches would save a lot of lives if seriously applied to medical trials. Would get rid of existing placebo trials that kill people.)

 

...This way I could correctly predict about 80% of the tags from the text alone.

I wanted to comment that is a surprisingly good result.

Why? Because most TW are authored in the secrecy of one's own attic. TW is not written in a networked system with any referent lingo on tagging ... when you author its "me, myself and I" ... we have no "auto-wizards" saying "You perhaps don't mean Tagg, but Tag?"

So 80% on the ball is pretty amazing IMO. **I think that is worth noting**.

 

The problem was that, to me, many of the tags were meaningless and were used internally to organise the TW.

Right. Partly its a mediation of "private language" ... I DO  create tags like "miniFrugal" that I know what *I* mean to myself but anyone else would struggle with ... that would need "translation". BUT, I never thought you were interested enough it would go shareable public ... :-) Partly (and often wholly) tags are content organisers, not semantic labels.

In a second experiment I totally ignored the assigned tags, and predicted the tags from

a TF*IDF analysis of the text. This made tags that made much more sence to me, but the

predicted tags often missed the supplied tags.

That is interesting. I suspect part of that result may devolve to the fact that wiki "made in your own attic" will differ on *tags* than a wiki made in "served networks" where commune lingo may get more attention -- just an hypothesis.

 

In my opinion the TF*IDF were better than the assigned tags since they had nothing

to do with the organisation, but more to do with the actual words in the text.

 Personally I like idea one derives "semantic heft" directly from units (tiddlers), rather than from labels of them. For two reasons (1) the less I have to do to add manual tags the better; (2) I know there are patterns I don't see that smart code likely can.

But, at the same time, any TW tag is a "label applied" to a tiddler -- a distance between the tiddler and its manifest content.

FYI I'm a big fan of Twiitter where #hashtags are always inline. No separation of content from organization. Its a neat approach on content cognisance. Twitter is maybe extreme in its #hashtaggery but its effective in terms of finding stuff well enough. But, of course, Twitter usage of #hashtags is purely about flagging content, whilst in TW tags do several jobs.

YES :-)  -- Given my earlier observations, perhapse we could distinguish two types of

tags. The #inlineHashTags could have something to do with the content of the containing paragraph. The tiddler tags could mean "tags used to internally organise the TW itself"

Just FYI, at the moment TW does not support out-of-the-box inline taggery, only the label type.

Best wishes

Josiah

[TonyM]

To all,

I love this Kind of conversation because we are peeling back the complexity in the universe, tiddlywiki and our minds. I would like to suggest something which I think many of you may have missed.

Tags can be tiddlers and thus can also be tagged

It is a simple matter of tagging your tag tiddler,s to create a namespace, for example (not something I do much anymore) if you had a set of tiddlers that would be subsequently used as tags to represent the status of a tiddler, you can tag "new" "wip" and "completed" tag/tiddlers with status. When scanning the tags on any tiddler with little effort you can interrogate the tags for belonging to the set of status tags. For each tag test if the tags tiddler has a tag of status.

Making a tiddler for each tag also allows you to define or document more about that tag.

The reason I no longer do this is tags are not as versatile as a field, and when something changes status, it can be helpful to know when it happened. So I now follow these rules.

• If a tiddler is a task tiddler-type field = task then it can or will have a field called item-started, 
• If item-started is blank then it is a new task, to indicate it is started I put a time stamp in item-started. 
• If something is started it is now "work in progress" and can now have the option to item-completed or item-cancelled shown.
• In most cases I need only test if a field has a value, and in others or sorting I may use its date/timestamp to order, or predict a due date etc....
• This is using fields as if they were tags, yet these tags can "not exist", exist without a value, exist with a value, that value can provide a relative time relationship.

If you really like the ease of setting and removing tags you can also use Mario's alt-tags to have multiple tag fields. Such tag fields can be used as subject or category organisation 

If you want one value only assigned for a given relationship, use a field that only accepts one value. Further if you want that value to be from a curated set provide a select statement with the possible choices or as I like to do use [has:field[fieldname]get[fieldname]]  to allow you to only select from existing field values.

Using TOC macros or the Kin operator I plan to for example to "tag" tiddlers that represent people, with a field called date-of-birth, if it exist it is a person, but when available can store the date of birth. I will then ensure the genealogical tree only concerns itself  with tags on tiddlers (eg parents) that are themselves people (having a data-of-birth) field. Thus the other tags remain free to indicate other qualities.

The simple act of ignoring tags prefixed $:/ is enough to keep system tags out of your models.

Finally another technique I use is a field than names fields. for example edit-fields contains a list of fields that I may wish to use on a given tiddler, these can be populated by button or template at creation time, and the order of these fields are displayed (which you can drag and Drop) to indicate the order, when a form is presented. Each field has a tiddler by its name, just as a tag can and this contains the definition, or edit macros to modify its contents along with the wikitext/macros to render the fields column heading and cell contents in a list/table.

I hope these inspire you or you can share or extend. 

The new subfilter operator allows you build logical filters such as work-in-progress = [tiddler-type[task]has[item-started]!has[item-completed]!has[item-cancelled]]

And use the filter "[all[current]subfilter[work-in-progress]]" or "[project[projectname]subfilter[work-in-progress]]" for all working progress items for the  projeectname

My belief is there is no complex system you can not model in tiddlywiki

Regards

Tony

[PMario]

On Thursday, December 13, 2018 at 7:19:16 AM UTC+1, TonyM wrote:

...

If you really like the ease of setting and removing tags you can also use Mario's alt-tags to have multiple tag fields. Such tag fields can be used as subject or category organisation 

I'm honoured. But it's Jed's stuff :)

-m

[PMario]

On Wednesday, December 12, 2018 at 5:54:41 PM UTC+1, Joe Armstrong wrote:

...

What I did was to use Baysian inference to "learn" the relationship between the words in the text and the supplied tags - so for each word in the text I caculate the probability that the tiddler has tag <T> (forall known tags <T>) - then in a second pass I tested the model and predicted the tags from the text. This way I could correctly predict about 80% of the tags from the text alone. The problem was that, to me, many of the tags were meaningless and were used internally to organise the TW.

That's interesting. ... But I think this has some "evolutionary" causes.

I think, it hasn't always been that way. Open classic.tiddlywiki.com ... You'll see wikipedia-like tag-box in every tiddler. ... The UI isn't "nice" with tags here. .. So a very common question in the group was: "How can I switch this box off?"

I personally prefer something that's called "TagglyTagging" (... oh we love those weird names :) I think TagglyTagging was introduced with MPTW (Monkey Pirate TiddlyWiki). For me it was a completely new way to work with tiddler titles. ... TT is a set of plugins, that allows you to visualize the relation between different tiddlers ... It speeds up navigation between related tiddlers, in a very convenient way. ... The "sitemap" view is what we call TOC (Table of Content) in TW5 now.

An other plugin, that imo influenced TW5 was: the fET-plugin (for each tiddler). It allows users to iterate over the tiddler store and create many types of "list-views". This plugin was highly influential for the TW5 list-widget, and <<list-links ...>> macros that we have today.

Both of those systems (mis)use tags to create internal structure, because the tagging mechanism was and is highly optimized. Both in the core-software and the UI. The core uses several caches to speed up tag and "backlink" lookups. ... We do have fields and filters, that are able to create invisible internal structure. But none of those possibilities offer the performance and "ease of use" from the UI perspective.

In my opinion the TF*IDF were better than the assigned tags since they had nothing

to do with the organisation, but more to do with the actual words in the text.

For me it would be very interesting to have a mechanism, that would suggest "meaningful tags" by analysing the prose text(s). ... but it needs to work without a 3rd party server. It should be integrated into TW. ... Is this possible?

have  fun!

mario

[Alex Hough]

Mario,

" a mechanism, that would suggest "meaningful tags" by analysing the prose text(s). ... but it needs to work without a 3rd party server. It should be integrated into TW. ... Is this possible?"

Wordnet [1] might be useful. 

WordNet® is a large lexical database of English. Nouns, verbs, adjectives and adverbs are grouped into sets of cognitive synonyms (synsets), each expressing a distinct concept. Synsets are interlinked by means of conceptual-semantic and lexical relations. The resulting network of meaningfully related words and concepts can be navigated with the browser. WordNet is also freely and publicly available for download. WordNet's structure makes it a useful tool for computational linguistics and natural language processing.

I use it a lot to find words at higher or lower levels of abstraction. It would be wonderful if, when tagging, words from Wordnet were suggested...

Alex

[1] https://wordnet.princeton.edu/

--

You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.

To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.

To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at https://groups.google.com/group/tiddlywiki.

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/90d0b5a8-88b2-4489-ac3d-ed724f08b01c%40googlegroups.com.

[@TiddlyTweeter]

Alex & PMario

External reference could be useful. Personally I prefer associative visual maping like VisualTheasaurus.

BUT before that, surely, we need to get somewhat more COMMUNAL on taggery?

I mean, why look up Godzilla when simple sharing could do it?

J.

[@TiddlyTweeter]

PMario wrote:

For me it would be very interesting to have a mechanism, that would suggest "meaningful tags" by analysing the prose text(s). ... but it needs to work without a 3rd party server. It should be integrated into TW. ... Is this possible?

 Great query. Internally would need limiting ... associative broader remit (all homonyms, synonyms etc)  library size too vast?

Smart text analysis possible (stemming; redundancy reduction) but limited.

IMO this could work for "domains of interest". For instance, the obvious examples would be "tags for components of TW" which could be a data tiddler listing commonly used synonyms. Obvious would be stuff like "plugin" v. "plugins"

I doubt internal "all and every possible" is possible without external referent input. BUT I doubt its needed to assert a reliable tag. Taggery is in "domains of interest" and that finitude looks good enough?

Just thoughts

Josiah

[@TiddlyTweeter]

Alex

Is it in many languages?

J, x

On Friday, 14 December 2018 13:39:23 UTC+1, AlexHough wrote:

[Alex Hough]

https://babelnet.org/ == many languages

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/a2ab39f0-0569-4cb1-90fb-a7bb2303c36a%40googlegroups.com.

[Marcel Otto]

Am Dienstag, 11. Dezember 2018 19:03:30 UTC+1 schrieb Joe Armstrong:

Thinking out loud here ...

I've been thinking more about tags. One problem is that tags are rather vague and are written in different human languages.

One way out of this might be to adopt the wikidata word definitions. For example, I am, unambiguously

    https://www.wikidata.org/wiki/Q1691321

There are actually several Joe Armstrong's (for example, https://www.wikidata.org/wiki/Q712592)

These Q numbers uniquely define subjects and objects. Verbs (or predicates) are given by P numbers

so https://www.wikidata.org/wiki/Property:P178 means "the organisation or person who developed the item.

I really like this idea. The natural-language tags of a TiddlyWiki could be linked to global URIs (like the ones from Wikidata) by adding a context to a TiddlyWiki in the same way JSON-LD adds a @context to JSON documents to give globally understandable meaning to application-specific property identifiers: https://en.wikipedia.org/wiki/JSON-LD

 

in RDF speak the triple

    {https://www.wikidata.org/wiki/Q1144644, 

      https://www.wikidata.org/wiki/Property:P178, 

      https://www.wikidata.org/wiki/Q17031730}

(BTW I recommend clicking on these links and playing around - there's lots of interesting

data in RDF tuples and the above links are a good place to start looking)

Means "TiddlyWiki developer Jeremy Rushton"

These triples encode facts in a hopefully reasonably clear manner.

So now the N$ question - can we automatically analyse a tiddler and turn it into a set

of RDF tuples. If we could then we could add these to the huge databases of RDF tuples

and possible find stuff in a clever way.

That would be a very ambitious endeavor as it would require solutions to two very hard problems: 

1. The long-standing research problem of entity recognition and linkage: https://www.stardog.com/blog/entity-linking-in-the-knowledge-graph/

2. Connecting the recognized entities semantically properly. For example, how would you detect from the sentence "Jeremy Ruston started the TiddlyWiki project in 2004." to use the https://www.wikidata.org/wiki/Property:P178 property?

Regarding the first problem, which would already solve your initial tag ambiguity problem, research has led to quite some progress. This library, for example, offers a solution in JavaScript: https://github.com/spencermountain/compromise

The filter notation in the tiddlywiki reminds me very much of prolog, and I guess with a but of

work SPARQL queries might be possible (SPARQL is an RDF query language)

Even without a solution to the mentioned second problem (allowing to generate proper RDF triples like the one you mentioned), we could already do interesting SPARQL queries with recognized Wikidata entities for content tags. We could for example query for all tiddlers tagged with a computer scientist and would get the tiddlers tagged with "Jeremy Ruston". A SPARQL engine for in-memory data in JavaScript can be found here: https://github.com/antoniogarrote/rdfstore-js (I can't resist to also mention that a SPARQL engine also exists for the BEAM: https://github.com/marcelotto/sparql-ex ;-))

Cheers

/Joe

Cheers,

Marcel 

[Alex Hough]

 further thoughts from Wordnet....

the tag needs to classify the contents of the tiddler. So for example Concepts [1] classifies many tiddlers as belonging to the class "Concept"

Wordnet has hierarchies of meaning based on synsets.

suggestions for choosing words higher and lower in hierarchy would be useful for me. I am forever going to Wordnet and choosing the right word. Integration would be good

Alex

[1] https://tiddlywiki.com/prerelease/#Concepts

--

You received this message because you are subscribed to the Google Groups "TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+...@googlegroups.com.
To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at https://groups.google.com/group/tiddlywiki.

To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/091b83d3-9a94-4010-bb2a-e4feca649709%40googlegroups.com.

[@TiddlyTweeter]

Ciao Alex

Part of the problem here is you are drifting back to TAXONOMY. Its problemmatic at scale. Though within a delimited field of interest Taxonomies can work pretty well.

As far as I understand it the most productive flexi approaches have been "natural language theasurus'" combined with FOLKSONOMY. 

Josiah

[Thomas Elmiger]

Hi Mario and all,

The most useful suggestions for tags for me come frome other tiddlers in a list – if they are in the same list, there could be a relation between tiddlers I did not see before ... so I built something I call "context tagging" or "list-based tagging" for my listreveal plugin. This solution makes it easy to apply tags used in other tiddlers in the same list.

See entry 03 in the list on the readme tab here: https://tid.li/tw5/plugins.html#%24%3A%2Fplugins%2Ftelmiger%2Flistreveal

Have a nice weekend,

Thomas