Tags:
classification2Add my vote for this tag discussion2Add my vote for this tag findability1Add my vote for this tag navigation1Add my vote for this tag tagging4Add my vote for this tag taxonomy1Add my vote for this tag twiki_community2Add my vote for this tag create new tag
, view all tags

Discussions on Content Tagging of TWiki.org Topics

Lets use this topic to discuss tagging of TWiki.org topics.

FYI, I am monitoring the tagging activity on a daily basis to fix/remove tags. Evaluators try out tagging and leave inaccurate traces behind.

-- PeterThoeny - 05 Apr 2006

Copied from deleted TagMeTalkAbout_tagging topic:

Here's an idea: have a conversation about every tag... makes me think tags should have been wikiwords... perhaps tag topics are lower case and have a VIEW_TEMPLATE that includes the TagMeSearch topic...

Thoughts?

-- MartinCleaver - 19 Mar 2006

Tag:extract_stuff

What is this tag meant to mean?

-- MartinCleaver - 05 Apr 2006

Not a clue, but are we going to have a TagMeTalkAbout_ for all of them? This isn't a jab at you: I'm just wondering if this tagging stuff is going to wind up creating yet more topics in Codev.

-- MeredithLesly - 05 Apr 2006

I was wondering too. I think Rafael added it as a reminder to extract useful stuff to other topics.

-- PeterThoeny - 05 Apr 2006

I replaced the TagMeTalkAbout_<tag> link in TagMeSearch with a link pointing to this topic here. I think one topic per tag is a bit overkill. We can revisit this if we see that this topic here is growing too much.

-- PeterThoeny - 05 Apr 2006

I added TagMeTalkAbout_<tag>

Well why are tags not wiki words? I wanted to talk about the ExtractStuff tag, not a mushed together discussion about tags in general.

-- MartinCleaver - 05 Apr 2006

Most tags are singletons, so WikiWords would help just for compounds.

If we see lots of activity in this topic we can go back to the one-topic-per-tag approach. However, I am concerned about clutter if we do that.

-- PeterThoeny - 05 Apr 2006

Martin: point noted and reasonable.

My first impression on this mechanism is that it's a good idea that isn't working very well. The largest items in the tag cloud have somewhere between 45-50 topics associated with them, some of them with dubious connections. I've also noted that many of the (at least transitorily) active discussions have no tags at all, either from the creator or any of the participants.

I should qualify that somewhat: Some people may have found personal tags useful for them to find topics they're interested in. I have no way of knowing that. As a mechanism for organising content, however, this would appear to be a failed experiment so far.

-- MeredithLesly - 05 Apr 2006

Peter: yes, that's the idea of extract_stuff. When I'm browsing quickly a topic and found something that should/could be extracted i just tag it and continue browsing the next topic. Later on I would go back to the extract_stuff marked topics and extract the stuff.

OTOH, even as tags are a good idea, they don't work well if the comunity is not too big or don't have a well-defined language for tagging. Let me elaborate.

If you allows "free tagging", that is, people can create whatever tag they want and tag topics accordingly, if the universe of taggers is too small (and I guess "too small" means less than 100) then you may end up with several tags for the same meaning, and with the same tag having different meaning to different people. If the universe of taggers is big enough (I'm thinking del.icio.us here) theh the "shared language" of each subgroup will tend to stabilize, so people in the same subgroup will be able to find content tagged by others in the same subgroup. (If someone can understand what I'm saying and can reparaphrase it, be my guest).

OTOH, having a fixed set of "tags" with well-defined semantic allows for easier tagging and finding because everybody will speak the same language.

So, I think that the best we can do "now" is to take whatever list of tags we have, try to extract and consolidate them in a rational list with a definition for each one, and prevent people from creating new tags (perhaps only allow core to create new tags). Also, the TagMePlugin could be modified to have "public" tags (everibody see them) and "privates" tags (only the user see them). From time to time, there can be a review of "private" tags to look for new candidates for the public ones.

-- RafaelAlvarez - 05 Apr 2006

Hm, aren't there already well designed lists of 'tags' used successfully since years by librarians? Why reinventing instead reusing the obvious? Just my 2c.

-- FranzJosefSilli - 05 Apr 2006

Silly boy. NIH.

-- MeredithLesly - 05 Apr 2006

Franz, to be honest I dind't know. Can you give a pointer?

-- RafaelAlvarez - 05 Apr 2006

There is no such thing as the list of tags that can be used universally. The classification systems developed by librarians (Dewey Decimal Classification, or the derived Universal Decimal Classification by Paul Otlet) reflected existing book collections and where designed to distribute the available books evenly over the libary. It works by dividing the existing books in broad subject categories (see also Wikipedia:Dewey_Decimal_Classification and Wikipedia:Universal_Decimal_Classification). Notorious is the DDC's skewed bias towards Christianity versus other world religions. DDC and UDC are more or less 'frozen': changes do occur very seldom.

A better system to descibe the content of works is the Library of Congress Subject Headings (LCSH) (Wikipedia:Library_of_Congress_Subject_Headings). This is a strictly controlled vocabulary, for and by librarians. Many familiar terms are not listed or outdated (read the amusing paragraph on "Afro-Americans").
LCSH terms are made for consitency. For example 'body temperature regulation' is used in place of 'thermoregulation'. If end users are not familiar with LCSH, they may incorrectly assume their library has no items on their desired topic, if they chose to search by 'subject' field, and the terms they entered do not strictly conform to a LCSH. For example 'body temperature regulation' is used in place of 'thermoregulation'. Thus the easiest way to find and use LCSH is to start with a 'keyword' search and then look at the Subject Headings of a relevant item to locate other related material. (citing Wikipedia)

LCSH is often not specific enough. Despite LCSH's wide-ranging and comprehensive scope, there are libraries where the use of LCSH is not ideal or effective. To deal with these types of collections and user communities, other subject headings may be required. The United States National Library of Medicine developed Medical Subject Headings (MeSH) to use on its many health science databases and collection. Many university libraries may apply both LCSH and MeSH headings to items. In Canada, the National Library of Canada worked with LCSH representatives to create a complementary set of Canadian Subject Headings (CSH) to access and express the topic content of documents on Canada and Canadian topics. (citing Wikipedia)

It is a very normal situation that subject fields (and even companies) have their own classification schemes and methods.

Summarizing:

  • Traditional classification schemes reflect collections, not knowledge
  • They are culturally and era biased
  • The used terms do not reflect natural language or modern culture
  • Terms are designed for consistency and control, not for end users
  • The classification systems hardly change

Now comes tagging, made famous by "folksonomies": (also quoted in FolksonomiesClassification): The folksonomy is a means for people to tag objects (web pages, photos, videos, podcasts, etc., essentially anything that is internet addressable) using their own vocabulary so that it is easy for them to refind that information again. The folksonomy is most often also social so that others that use the same vocabulary will be able to find the object as well. It is important to note that folksonomies work best when the tags used to describe objects are in the common vocabulary and not what a person perceives others will call it (the tool works like no other for personal information management of information on the web, but is also shared with the world to help others find the information). http://www.vanderwal.net/random/entrysel.php?blog=1635

Tagging was met by skepsis (translation: skepticism) by librarians, but the line is not clear cut now. For instance, http://www.librarian.net/ has a tag cloud (at the bottom). And read the conclusion of the research paper http://www.adammathes.com/academic/computer-mediated-communication/folksonomies.html:

A folksonomy represents simultaneously some of the best and worst in the organization of information. Its uncontrolled nature is fundamentally chaotic, suffers from problems of imprecision and ambiguity that well developed controlled vocabularies and name authorities effectively ameliorate. Conversely, systems employing free-?form tagging that are encouraging users to organize information in their own ways are supremely responsive to user needs and vocabularies, and involve the users of information actively in the organizational system. Overall, transforming the creation of explicit metadata for resources from an isolated, professional activity into a shared, communicative activity by users is an important development that should be explored and considered for future systems development.

This article lists many cons and pros of tagging, and even to summarize these would get too lengthy here. I invite everyone interested in tagging to read the article.

My take on TWiki tagging
Currently the goal of tags on twiki.org (or in fact the Tags plugin) are not clear. I see two diverging options: tagging or classification. We must decide which path to follow. Either:

  1. Use full blown tagging. Use the perspective of the individual: let everyone add tags as they think fit. Don't remove tags. Use personal lists and voting. Allow personal tags. Many delicious users coin a website with the tag "toread"; this has a meaning for these individual users.
  2. Use tagging as an easy to add-and-edit classification. Try to keep the tag list as small as possible by removing duplicates and seemingly senseless tags.
While I can see use for both ways of tagging, I think twiki.org (and most intranets) are helped more by option 2. Why?
  • The subject scope of the site is known (or at least its boundaries can be defined by a certain set of tags - not meaning the tags cannot change or be added to)
  • The number of users is limited (while there are many on twiki.org, the visitors here are dwarfed by the amounts delicious gets)
    • This means that you don't get the benefits of "cultural tagging": each culture that defines its own tags for a certain subject ("movies" or "cinema" or "film" for example).
    • So one topic will get only a few tags or perhaps even one or none. With less tags, it becomes more important that the "right" tags are used.
      • With no control and just a few tags, the searcher's search tags have less chance to match the tagger's describing tags.
    • We want findability: if we are looking for topics on a given subject, we want all topics on that subject. That requires consistency
    • Consistency means: less use for individual perspective. So no toread tags for instance, or extract_stuff.
    • It also means: someone looks after the tags, and predicts visitor's search patterns, correcting tags if they don't fit in.

-- ArthurClemens - 05 Apr 2006

At our recently installed intranet TWiki tags are working really well as a means of classification. The voting mechanism is not understood by our users.

-- ArthurClemens - 05 Apr 2006

Hm, maybe the JSC for Revision of AACR has also some pointers.

-- FranzJosefSilli - 06 Apr 2006

MichaelSparks and I discussed this at considerable length a couple of years ago. Raf, Martin, IIRC know you were there at the time! Our conclusion was that the only important perspective is the perspective of the individual. Within any group, each individual will have different mental vocabularies and taxonomies, and each subgroup within a larger group will have different vocabularies and taxonomies. It is possible to impose a taxonomy through cultural engineering; by imposing from above, or by de facto. However such taxonomies are extremely brittle, and inevitably end up being more of a burden than a solution.

"Full blown" tagging as described above is a more democratic means of cultural engineering; but it is still a process of imposing the will of the majority on the minority. What we lack is the ability to support individual taxonomies keyed to the world views of individuals and groups. For example, the "tagging" of Codev is currently being done by people familiar with TWiki, and more importantly, mainly by developers. So the world view being imposed is not useful from the perspective of a newbie, and not useful from the perspective of a quick search for a cookbook.

That's not to say tagging doesn't have value; it does, it has tremendous value. But please, don't think that tagging is a silver bullet that will solve all the taxonomy/vocabulary problems. It's far too crude for that. The main benefit of tagging I see in Codev at the moment is this: Codev has been around for a long time, and there are several intersecting indexes in place already (categories, forms, classification). Each of these is in some way unmaintained and therefore stale. Tagging is an opportunity to update our vocabularies with more modern terms. However my gut tells me that like all the other indexing schemes, without constant TLC it will also grow stale.

So, what's the answer? Well, individual tagging is essential. But that doesn't help newbies, or occasional visitors. So how about applying your brainpower to classifying individuals and then collating their tagging activity to build statistics-driven indices for the groups they belong to? For example, I am a developer. Each time I tag something, that needs to add weight to the tagged object for the developer SIG (special interest group). I am also a keen cyclist, so it also needs to add weight to the cycling SIG. And a classical guitarist. etc. Ultimately (assuming enough people engage with tagging) any guitar-playing, cycling developer has a ready-made special-interests index. Oh, and at the same time, in a blinding bit of accidental social engineering, you have an extremely powerful, self maintained, directed-marketing database.

Until tagging has some collation capability like this it is, IMHO, just another index.

-- CrawfordCurrie - 06 Apr 2006

And a very poor index unless a sufficient percentage of users are disciplined enough to tag topics or the ability to find untagged-topics is created and someone(s) willing to do the tagging.

Note: at last count there were 12 tags assigned to this topic. This demonstrates the current display problem that results from too many tags assigned to a topic.

Someone (sorry not to credit you) suggested that we need tabs ala Wikipedia: one for the topic and one for the comments. The tags on this particular topic (which of course may be unusual) suggest tha t there should be a third tab for "index" type stuff if we were to go that way.

-- MeredithLesly - 06 Apr 2006

For people to tag, they must be reminded about the tag cloud. For this reason I think showing the tag cloud at all times is vital. How much of it is shown is another matter.

-- MartinCleaver - 06 Apr 2006

I'd like to pick up on Crawfords suggestion that "...the only important perspective is the perspective of the individual." I don't buy this proposition and believe that it reflects a very limited view of how groups work. It suggest, in essense, that the group's understanding can be no more than the sum of our individual perspectives. A wealth of social research suggest that one of the most basic processes of becoming a group is negotiating shared meaning. If one has worked with a team, you know that the first step in coming to shared understanding around basic vocabulary.

This suggest to me that in using a wiki, the groups negotiation and agreement about the basic taxomomy of their work is an critical step to creating a body shared knowledge. This process can not be automated or reduced to a coded solution. It requires people actually striving to understand each other. Yes, the resulting taxonomy may be "brittle" but then group cohesion is a rather delicate thing that must be attended to and nurtured. I find the suggestion that the only way this can be done is through "social engineering" is a quite cynical view of group process.

Coming back to a more practical discussion and referring back to Arthur's analysis (which I fully agree with), to me, the second type of classification system is essential as it provides the basic navigation structure for a site. The first type of classification (folkonomies) may be useful in some contexts but is not essential. The proposition that the first should be completely thrown out is, to me, absurd.

In my experience helping groups set up their basic taxonomy using TopicClassificationAddOn, the process of their deciding how to organize their content was valuable in itself for helping them come to grips with what they up to. And the resulting navigation structure so produced, while modest, served it purpose. For all the talk of coding some perfect tagging system to replace this process, I have yet to see anything that comes close to providing this modest, but essential, function.

I'd like to add one last comment about this whole discussion as it relates to TWiki.org. This site is probably not the best measure of a classification system because it is, I believe, an atypical group context. It reflects, at best, a fragmented and transitory community lacking basic structures of accountability trying to grapple with a fundamentally difficult task - that is, designing software who's purpose is an evolving target. (BTW, that's why I enjoy hanging out here. smile ) But this is not, I would suggest, typical of most groups that would wish to use TWiki. Furthermore, I would suggest that the failure of past and current efforts to organize TWiki.org is not a failure of technology (nor of our collective ability to code a perfect system), but rather a failure of organization and group culture. I don't mean to blame anyone or disparage our efforts by saying this! I'm just suggesting that just because we have not been able to create a satisfactory classification/navigation system should not be taken as proof for the failure of existing mechanisms (e.g. TWikiForms) nor the validation of some yet-to-be-invented mechanism.

-- LynnwoodBrown - 06 Apr 2006

Sorry, Lynnwood, I didn't make myself clear. I meant the only important perspective to automation. You are absolutely right in your comments about shared vocabularies, on the building of organisations. However the subject of our original discussion, and my comments above, was the spontaneous emergence of organisation within large, constantly changing, groups, and how to enable it. I think that's where tagging, and tag clouds in particular, are aimed. I don't think any of the existing models - including tagging - fail per se, I just think they generate static views that require high maintenance, and that's why they tend to go stale.

-- CrawfordCurrie - 06 Apr 2006

There is also (perhaps) an assumption that a group consensus (probably not with 100% agreement, but that's the nature of humans) is allowed to emerge and is not overridden by a higher authority (and I don't mean god). And yet without pruning, so many tags emerge that one gives up on tagging at all. Thesis, antithesis...where is the synthesis?

-- MeredithLesly - 06 Apr 2006

There should be a search for 'no-tags' else we will not know what has not had coverage.

-- MartinCleaver - 08 Apr 2006

Is it possible to browse twiki tagged items by more than one tag, for example view all items

tagged with both "tagging" and "intranet" ?

(Like on http://del.icio.us)

-- RebeccaCox - 10 Apr 2006

Rebecca, I've added your question to TagMePluginDev. -- ArthurClemens - 10 Apr 2006

I thought the idea behind tagging on twiki.org was that it was self-organising and all that, not to mention being in part intended for personal use. Personal tags might or might not be useful for other people but to remove them because they don't meet a monitor's standards defeats the very purpose of folksonomy.

-- MeredithLesly - 04 May 2006

Folksonomy tagging is like blogging, something for an individual. And then with some glue to link things together. Wiki tagging is different, something for the team. The tags are shared and "owned" by the community, not unlike wiki content. Tags applied by all are shown on a topic. This is different from Folksonomy tagging.

Granted, there are better answers to one person monitoring the wiki tagging activities. The tool could be enhanced so that the community can police itself, aka the wiki way. Also, public wikis are a special case; monitoring wiki tagging is not really needed in a wiki at the workplace.

Someone has to monitor the tagging activity on a public site, or it would quickly result in useless data. Consider these tags recently applied on TWiki.org by people just poking around the site: User 1 creates new tags 'shasha' and 'wawals', user 2 tags TWiki.WebHome with 'archive_me', user 3 tags Main.WebSearch with 'deployment', and Sandbox.WebHome with 'version_control', user 4 tags Codev.WebHome with 'windows'. This all within the last 24 hours. I am cleaning up all that stuff. Naturally I am also concerned if all tags of a user are deleted all of a sudden because it removes tags from topics that are considered 'done' from a tagging perspective.

So overall, folksonomy is useful to find related content quickly, provided that the tagging activity is monitored to ensure the quality.

-- PeterThoeny - 04 May 2006

Peter I'm concerned that your managing this will take time away from other more productive work. A tagging system shouldn't need that kind of monitoring. Consider that del.icio.us has asdf tags dating back to 2004.

The tagging system should be automatically doing the filtering of irrelevant tags. Maybe it should not display tags on a topic until there are more than two votes for the tag.

The mechanics of TWiki's tagging system aren't different than any other tagging system. We can't expect people to use it any differently.

I'm not sure I that there are two different kinds of tagging, individual and community. It's more like one is an emergent property of the other, i.e. the "community tags" that are "owned" by the community are really only the private tags that coincide most frequently. You can't have one without the other.

-- SamHasler - 04 May 2006

And I'm not sure I agree with that, at least with in the case of a small community. It's hard for properties to emerge with so few data points.

OTOH, Peter's intent (if I've understood it correctly) to use "tagging" as a means of classifying topics makes sense to me and I've posted something about that in DocumentationFocusGroup.

-- MeredithLesly - 04 May 2006

I've nothing against it being used for some coordinated classifying effort. I don't think that conflicts with using it also as a private tagging system. The votes of the community/coordinated tagging should be enough to drown out the private tagging as background noise.

I've even started to use it as a way of gauging the how easy different installs are to perform/follow by making requests to add the easy_install tag with convenient tagging links to IndigoPerlCookbook and FullyWindowsNativeInstallation.

-- SamHasler - 04 May 2006

I don't think that using the same list of tags for both purposes will work, I'm afraid. For classification purposes, you want a relatively small number of tags; otherwise it's too dificult to choose from the plethora of tags. And, as you noted, it's not the best use of Peter's time to have him sorting through and consolidating tags, not to mention that the very act of managing/merging/deleting tags works against using tags for personal reasons.

-- MeredithLesly - 05 May 2006

Currently, the tag system seems to be used mostly for classification. Topic tags are added by users only when missing rather than as a popular vote. I think this is good since spending effort retagging already classified topics is not very helpful.

However, the idea of using tags as a personal marker of interest, is a good one. This is quick and convenient.

Merging the two purposes has an advantage of encouraging users to spend the classification efforts, but it is not as obvious or intuitive. I'd like the ability to both provide classification tagging for the community and to add personal tagging for quick reference.

I'd like to retain the ability to tag something as "Interesting" or "ToBeRead" without necessarily determining the exact classification. I'd also like to be able to given a classification WITHOUT making the topic one of personal interest.

Perhaps the ability to create "private" tags (or least some defined as personal only, like mytodo, mywatch, myread, immediatenotify) would help separate these two purposes more usefully. In addition, a distinct set of "personal" tags could be used for notification control as well (immediatenotify and webnotify could use these for topic lists).

-- JadeCravy - 17 Jun 2006

Here's what I want in a tagging system:

1) I want to have a configure option, (similar to ALLOWTOPICCHANGE and friends) that preferentially allows only certain users (or groups) to add to the global tag set. Or anyone at all on a wide-open site, of course.

2) I want any user to be able to create and store private tags, (probably stored in their "home" area) that don't necessarily interact with the global set.

3) I want to be able to set tags on either a site-wide or a web by web basis, analogous to TWikiPreferences vs WebPreferences.

4) I would love, in the future, to have "subcategories" (tag hierarchies). A simple alphabetical list become cumbersome after a few dozen entries. We need a way to clasify the classificatiions. smile

I agree with ArthurClemens regarding the value of tagging for "easy to add-and-edit classification... [keeping] the tag list as small as possible" This is exactly what I want!

Where I am currently working, I've requested te installation of TagMe Plugin. An engineer and I had a "discussion" about that today. He wants the company (and therefore, me) to use only an internal delic.iou.us clone for tagging and bookmarking.

I see this as overkill, messy, and not what I'm looking for. Reading over the discussion has given me a better idea of what I do want and more reasons why the "everyone in the company tags anything, anywhere, with any words they make up" is not going to meet my goals of classifying and categorizing our docs. (Said bookmarking system currently has many duplicate entries, as well as nearly identical tags such as ops, operations, mediaops, and media-ops. The more I look at TagMePlugin the more I like its design goals.

Thank you Peter, Arthur, and everyone for your thoughtful comments above. I plan to reference this page in my future arguments.

-- VickiBrown - 23 Sep 2006

Interesting thought, an internal http://del.icio.us/ clone for the enterprise. Is it something you can buy /download today, or is it just "dream"-ware? smile

-- SteffenPoulsen - 23 Sep 2006

Vicki - To me, you're describing two different classification systems: one that is more tightly managed by a few people ("top-down" if you will) to provide the basic navigation structure for the site and the other a completely open, user generated ("bottom-up") system. TagMePlugin seems better suited to the second use. The first is what I had in mind when first creating TopicClassificationAddOn. I'm in the process of developing a new version that will incorporate many of the features you list such as being customized to individual webs and hierarchical. Ideally, it will incorporate some kind of FacetedNavigation framework.

-- LynnwoodBrown - 23 Sep 2006

Lynnwood - Interesting. I would have thought TagMe was more suited to the first system you describe (certianly more so than the "del.icio.us clone".

I'll go read up on TopicClassificationAddOn (and start a new "discussion" weith the engineer where I work?? smile

Steffen - It's not dream-ware (though it's pretty buggy)... Whether it will ever be released as a product, I couldn't begin to guess.

-- VickiBrown - 25 Sep 2006

I think I understand how TagMe works for topics, but how can I create a taxonomy for file attachments? I'd like to add tags for a searchable document management web.

-- GeneHiggins - 28 Mar 2007

A very interesting topic to read.

As Arthur pointed out, there seem to be at least two main approaches to tagging:

  • A full blown tagging approach
  • keep the tag list as small as possible approach

I also agree very much to Crawford's words on the fruitlessness of imposing a taxonomy, and also to the view that out of a full blown system spontaneous order often emerges. Also, IMHO, receiving an email from Peter that some tags were removed as they were not deemed appropriate, is counter productive to all involved.

I think that the two approaches might be reconciled, if the current tagging system would make a distinction between three kinds of tags:

  1. Private tags which serve as a personal bookmark or a TODO action (such as Interesting and ToBeRead that were mentioned above), and were created specifically for that purpose in mind.
  2. As yet a private tag. A tag recently created with the purpose of it becoming a public tag.
  3. Public tagss

So how does the above distinction help?

Tags from the first group will be shown (both in tag clouds and in the header of topics) only to the user who created them. my ToBeRead pointers will also differ from another user ToBeRead pointers (if he has any). It is solely personal tag list.

Now, how does one distinguish tags from the second group from those of the third group? by having a personal setting for each user as to what his personal VS. public threshold. For the more liberal ones of us, that threshold can be low as zero - show me (in tag cloud and header topic) all tags from group two - that way I can 'strengthen' new born tags by other users which seem to me appropriate and productive.

Others, who would like to keep the tag list as small as possible, would use a much higher threshold which suits their liking.

Another feature that would help to sieve the clutter out, is to have the tag search results list sorted according to the number of users who tagged a certain topic (hence also making the tagging system a voting mechanism). One might also consider the possibility that users would be given the option of decreasing a tag vote on a certain topic if they find that tag highly inappropriate .

-- EyalTeutsch - 04 Oct 2007

Interesting. I recall a talk with a person of the BBC News website one year ago. She described a somewhat similar system. The BBC had just moved from a classification system to a form of tagging to allow journalists more freedom, and to allow more specificity for local news (so that local events and people could be included as well. They system they used calculated the usage count of a given tag. If a tag occurred 2 or 3 times it was displayed publicly, otherwise it was only visible to the journalists.

As it is, TagMePlugin already can make the distinction between tags added by me and all tags.

-- ArthurClemens - 04 Oct 2007

I modified TagMePlugin.pm to include an action to show untagged topics. Quick and dirty:

--- TagMePlugin.pm.org	2010-04-29 09:29:18.000000000 +0200
+++ TagMePlugin.pm	2010-04-29 15:42:17.000000000 +0200
@@ -163,6 +163,9 @@
     elsif ( $action eq 'showalltags' ) {
         $text = _showAllTags($attr);
     }
+    elsif ( $action eq 'listuntagged' ) {
+        $text = _listUntagged($attr);
+    }
     elsif ( $action eq 'query' ) {
         $text = _queryTag($attr);
     }
@@ -553,6 +556,30 @@
 }
 
 # =========================
+# _listUntagged added by Henry, 2010-04-29
+# lists untagged Topics in current web, topics starting with 'Web' are omitted!
+sub _listUntagged {
+    my ($attr) = @_;
+
+    my $text="---+ Untagged Topics\n";
+
+    my @topics = TWiki::Func::getTopicList($web);
+    my $topic;
+    my $webTopic;
+
+    # get number of tags for each topic and display those with zero tags
+    foreach $topic (@topics) {
+        # omit topics starting with 'Web'
+	next if (substr($topic, 0, 3) =~ /Web/ );
+
+	$webTopic = "$web.$topic";
+	if (_readTagInfo($webTopic) == 0) {
+	    $text .= '   * [[' . $topic . "]]\n";
+	}
+    }
+    return $text;
+}
+# =========================
 sub _showAllTags {
     my ($attr) = @_;

-- HelmutEberharter - 2010-04-29

Thank you Helmut for contributing back to the TWiki community!

I invite you to get active with the community. I can request to have Subversion access for plugins. That way you can maintain the code directly in the repository. Details at Plugins.ReadmeFirst.

PS: It would be helpful to have docs on using the new feature.

-- PeterThoeny - 2010-04-29

Edit | Attach | Watch | Print version | History: r41 < r40 < r39 < r38 < r37 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r41 - 2010-05-13 - PeterThoeny
 
  • Learn about TWiki  
  • Download TWiki
This site is powered by the TWiki collaboration platform Powered by Perl Hosted by OICcam.com Ideas, requests, problems regarding TWiki? Send feedback. Ask community in the support forum.
Copyright © 1999-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.