Classification based on "keyword phrase"?

Hello - I have created a Classification Model using SVM that works well for a 3-class problem (Good, Bad, Neutral). The target is built into the training dataset and important keywords are discovered.
I've been trying to figure out if there is a way to build a Classification model that is based on keyword phrases (2 or more words) with Good, Bad or Neutral assigned to the "story".
For example, classifying news as being Good, Bad or Neutral for a company. The keyword phrases may be "failed to deliver" or "delivered on schedule" ... Each phrase is associated with Good, Bad or Neutral.
I don't believe this can be done with Support Vector Machine. Is there an alternative?
Thanks

The key hurdle to implement what you would like to do is how to create your "keyword phrases." If you have a way of doing it then you could use SVM to do the classification just the way you want it.
There are a couple of ways that you can achieve your goal. I will list some of them below:
1) Let's say that you have a way of extracting/creating your own set of "key phrases" either by hand or by some automated method. The next step would be how to create a classifier with this information. A simple approach is to consider the set of unique "key phrases" you've created as the attributes used for mining. Using your example, the phrases "failed to deliver" or "delivered on schedule" would be two different attributes. If a given document contained "failed to deliver" then we would set the value of this attribute to 1 otherwise to 0. In a sparse representation you don't need to represent the zeros, thus saving a lot of space. Now you can train SVM using this new set of attributes as inputs and your three classes as target. It is possible to create more sophisticated coding schemes that would take into account the number of times a phrase shows up in the document as well as the frequency of the phrase in the whole collection. But this initial scheme is a starting point that might help you see how to do it with ODM.
2) Using Oracle technology it is possible to automatically extract sentences or group of words to make new features. This would give you less control over the "key sentences" but may simplify the creation of "key sentences" for you. There are two approaches you can take. One uses Oracle Text feature extraction alone. The other combines the first approach with the NMF algorithm. Can elaborate further if this a path that you want to take.
Regards,
--Marcos

Similar Messages

Custom tag cloud webpart based on keywords from a Term Set in the Term Store .

Hi Guys
Does anyone have developed a custom Tag Cloud webpart in SharePoint 2010? I'm
looking for a custom Tag cloud based on keywords from a Term Set in the Term Store.
Unfortunately
out of the box Tag Cloud Web Part only reference the Social Tags from the User.
in my case editors have created pages and tagged a page using terms from the Term store.
What I would like to see is a web part which could display “Top Keywords”, and the output is something like this:
http://electronicdiscovery.info/what-additional-features-would-you-like-to-see-in-sharepoint-word-cloud-survey-electronic-discovery/
Thanks In advance
Tanaji Chavan

Hi Ramakrishnaraja ,
Thanks for sharing the links , i tried using web part in second link but that’s not working , have also gone through the code of that web part seems like it is referring to some list like
"Categories" and "Posts"
however I'm looking for a custom Tag cloud based on keywords from a Term Set in the Term Store .
Thanks

Using keyword "phrases" in podcast xml

Can you use keywords greater than one-word in the keywords section of the xml? For instance, if I wanted to use "science fiction" as a keyword, how exactly do I do that? I'm guessing it would look something like
film, "science fiction", review, criticism, "transformers revenge of the fallen", etc., etc.
Is this correct? Also, do the individual terms (such as "science fiction") count as one keyword or the number of words contained in the term?
Thanks for the help.

Nothing is said about phrases in the Apple Tech Specs: I doubt that using quotes is going to work. AFAIK you can only have single words (to a mximuim of 12) which should be separated by commas as you've done in your example.
Of course it you put in
science, fiction
and someone enters 'science fiction' in the search field then it will look for the two words and you stand a good chance of being found.

How do I allow firefox 4 to go onto a page based on keywords in address bar

Firefox 3.6 allowed me to use the address bar as a more clever way of browsing. I am not sure why this feature was removed but it used google to figure out from my keywords what website I meant.
e.g gamespot uk would bring up google.co.uk
speedtest bbmax would bring up speedtest.bbmax.co.uk
If anybody knows how to fix this it would be great otherwise I may jut downgrade.

Firefox 3.6 uses the Google "browse by name" feature, you can set Firefox 4 to use this feature by changing a preference.
# Type '''about:config''' into the location bar and press enter
# Accept the warning message that appears, you will be taken to a list of preferences
# Locate the preference '''keyword.enabled''' and make sure it is set to '''true''', if it is not you can double-click on it to change its value.
# Locate the preference '''keyword.URL''', double-click on it and change its value to the link show below
http://www.google.com/search?ie=UTF-8&oe=UTF-8&sourceid=navclient&gfns=1&q=
For more details on this see http://kb.mozillazine.org/Location_Bar_search

Sharepoint Online: Find an Item box is not finding all items in a list based on keyword search

I am using SharePoint Online.
I have a standard SharePoint list with a "Name" field (single line of text) and a "Description" field (multiple lines of text). The Find n Item search box is present as expected, and works very sporadically. For example, I
have an item with the word "white" in the Description field. Searching for "white" doesn't find the item. I have a number of items with the word "vending" in the Name field. Searching for vending returns some
of the items, but not all of them.
I have gone into the list settings (Advanced) and reindexed the list.
Here are two screenshots that illustrate the issue. The first is unfiltered. The second shows a filter on the word "vending". Note that some are found and some are not.
Filtered
Any idea what's going on?

I am also experiencing an issue with the find. If not the same issue David is having, I suspect it is related, so I will describe it here.
I created a list based on the Contacts content type and within a standard list view, entering values in the "Find an item" box will match content of some columns, but not others. The four columns that the find misses are Position(renamed from Job Title),
Email Address, Mobile Phone, and Years Of Service (a Numeric column I added to the list). I too have re-indexed the list and retested after several hours.
Full Name and Email Address are both native columns of the Contact content type and both are Single line of text. I am able to filter the list by finding partial matches on Full Name, but not Email Address.

Vendor classification based on its activity

Hi..
I have a requirement where i want to sort all the vendors based on its nature of activity. for e.g i have 10 vendors who provide only services and 15 vendor which supply office consumable, and 15 vendor which supply hig value material,
NOW if i need to have a list of vendors which provides only service how do i achive it.
is there any field in vendor master where i can specify the nature of activity.
Suggestion will be worth appreciation

go to ct04 and create a characteristic for lets say ur service vendor.dont use any table field.give the initial inputs in basic data tab.
in the value tab give all the vendor codes and description.in restriction tab give class type as 010.
now create a class thru cl02 for the characteristic with class type 010 and assign the vendor characteristic here.
now thru cl20n and class type 010 assign each vendors seperately with the created class and assign the objects here.
now whenever u use any vendor related reports or transaction use selection criteria as vendor by class,give the class name,class type and proceed.
regrards,
indranil

Changing Tax Classifications based on Departure Country Change

Hello Experts,
I am over writing 'Tax Departure Country' (LANDTX field) with selling country instead of delivering plant country in MV45AFZZ user exit.
Now...my requirement is that SAP should consider this new tax dept. country for 'Material and Customer Tax Classification' also. It doesnt take care of this new country automatically.
How can this be accomplished?
Thanks for your time and reply. It will be rewarded accordingly.
Sagar

Sagar Lipare wrote:>
>
>
> Now...my requirement is that SAP should consider this new tax dept. country for 'Material and Customer Tax Classification' also. It doesnt take care of this new country automatically.
>
Hi,
Do you want to consider Departure country as well as Tax classification in Material and Customer also? Now it consider only Departure country, not the Tax classification for Material and Customer. Is that what you mean by that?
If you want to have both you sales Departure country & Tax classification for Material and Customer, Please check the standard Access Sequence for Tax MWST. (Access Seq also MWST)
It has 3 condition table.
78 - Departure Country / Destination Country
2 - Domestic Taxes
11 - Export Taxes
If you can check table 11, you can see it include Tax classification for Material and Customer.
So what you can do is, create a condition table with you sales Departure country, TaxClass1-Cust. & Tax class. material.
You will be able to do what you require.
Best regards,
Anupa

Adding keywords and phrases for high search engine hits

I'm new to iWeb and most things seem to be working. I have one question and one problem, hopefully you guys can help out.
Question: how do I add html code that is not visible on the page but that will help Google and other search engines find my page using keywords, phrases, etc.? Please assume I know very little about this and need the holding hand explanation.
Problem: I have a media file portion on my website that displays pictures (working very well) and one that is supposed to show a movie (not working at all). The movie was made with iMovie and has an .m4v extension. When you click on the link for the movie it shows a blank screen on Firefox and a Quicktime logo with a big question mark on Safari. I'm attempting to view the website on my Mac with the latest Quicktime software on it but it doesn't work. Any ideas on how to fix it?
Some additional info: the hosting provider is 1and1.com and I have a Windows server package (unfortunately). Also, if I publish the site to a folder the movie area works flawlessly, it just does not work on the hosted site.
Thanks!!

You don't need to do post publication manual editing of your site to add keywords and other metadata. You can use the freeiWeb SEO Tool. If you're publishing to MobileMe you can add them after publication. If you're publishing to a commercial hosting server you'll need to publish to a folder on the desktop, run SEO and then use a 3rd party ftp client like the free Cyberduck to upload your site.
If you have to republish SEO remembers what you added and can quickly add them again before uploading.
OT

Keyword Based Smart Collections

I'm having some difficulty. Most of the time keywording is a great way to build smart collections. This time, however, I'm having difficulty. I had a second shooter at my last wedding. When we imported his images, they were all tagged with his name so that I could differentiate if I needed to. Trouble is, now my standard keywording technique is back-firing.
After building my smart collections for the different parts of the wedding based on keywords (e.g., smart collections for pre-ceremony, ceremony, details, reception and formals), I make one final smart collection that displays images that have no keywords. This way in case I failed to tag some images, I won't miss them. Trouble is, all of the second shooter's images do have keywords. They're tagged with his name.
So, what I need to do is build a smart collection that looks something like this:
Show me images only:
IF the image has NO keyword
OR
IF the image is ONLY tagged with THE PHOTOGRAPHER'S NAME
I've been tinkering with it for a while but there doesn't appear to be any obvious way to do this. Frankly it seems like pretty basic functionality. Even iTunes will do this kind of thing with Smart Playlists. So, can someone give me some direction here?

Set your smart collection to be
Keywords Is Empty
Keywords Contains "The Other Photogs Name"
And set the match parameters to "Any"

Build text-index based on a given list of words or phrases.

I'm somewhat of a beginner to this text-indexing. I've been able to build and query a simple text-index and even implement my own list of stop-words. However, I'd like to be able to control the set of words that are indexed.
For example, If I have a table that contains a CLOB field filled with text documents and I also have a list of 200 words:
"TUBERCULOSIS"
"DIABETES"
"CHEMOTHERAPY"
Can I generate an index that only indexes the words on that list and ignores all the other words? (the reverse of using a stop-list)
Also, could it be done with a list of phrases instead of single words:
"CARDIAC ABLATION"
"ATRIOVENTRICULAR NODE"
"PULMONARY ABSCESS"
Thanks.

Please see if you can use any of the pieces of the following example.
SCOTT@orcl_11gR2> -- table containing list of phrases:
SCOTT@orcl_11gR2> create table phrases
2    (phrase        varchar2 (21))
3 /
Table created.
SCOTT@orcl_11gR2> insert all
2 into phrases values ('TUBERCULOSIS')
3 into phrases values ('DIABETES')
4 into phrases values ('CHEMOTHERAPY')
5 into phrases values ('CARDIAC ABLATION')
6 into phrases values ('ATRIOVENTRICULAR NODE')
7 into phrases values ('PULMONARY ABSCESS')
8 select * from dual
9 /
6 rows created.
SCOTT@orcl_11gR2> -- ctxrule index on list of phrases:
SCOTT@orcl_11gR2> create index phrases_idx on phrases (phrase)
2 indextype is ctxsys.ctxrule
3 /
Index created.
SCOTT@orcl_11gR2> -- table to hold combination of documents and matching phrases:
SCOTT@orcl_11gR2> create table classifications
2    (document clob,
3      phrase       varchar2 (60))
4 /
Table created.
SCOTT@orcl_11gR2> -- context index on classifications table:
SCOTT@orcl_11gR2> create index class_phrase_idx
2 on classifications (phrase)
3 indextype is ctxsys.context
4 parameters ('sync (on commit)')
5 /
Index created.
SCOTT@orcl_11gR2> -- regular index on classifications table:
SCOTT@orcl_11gR2> create index class_phrase_idx2
2 on classifications (phrase)
3 /
Index created.
SCOTT@orcl_11gR2> -- table for documents:
SCOTT@orcl_11gR2> create table documents
2    (document     clob)
3 /
Table created.
SCOTT@orcl_11gR2> -- trigger to populate classifications table from documents table:
SCOTT@orcl_11gR2> create or replace trigger documents_bir
2    before insert on documents
3    for each row
4 begin
5    for r in
6       (select phrase
7        from      phrases
8        where matches (phrase, :new.document) > 0)
9    loop
10       insert into classifications (document, phrase) values
11         (:new.document, r.phrase);
12    end loop;
13 end documents_bir;
14 /
Trigger created.
SCOTT@orcl_11gR2> -- inserts into documents table:
SCOTT@orcl_11gR2> insert all
2 into documents values ('word1 tuberculosis word2')
3 into documents values ('word3 diabetes word4')
4 into documents values ('word5 chemotherapy word6')
5 into documents values ('word7 cardiac ablation word8')
6 into documents values ('word9 atrioventricular node word10')
7 into documents values ('word11 pulmonary abscess word12')
8 into documents values ('word13 word14 word15')
9 select * from dual
10 /
7 rows created.
SCOTT@orcl_11gR2> commit
2 /
Commit complete.
SCOTT@orcl_11gR2> -- resulting population of classifications table:
SCOTT@orcl_11gR2> column phrase   format a21
SCOTT@orcl_11gR2> column document format a34
SCOTT@orcl_11gR2> select phrase, document from classifications
2 /
PHRASE                DOCUMENT
TUBERCULOSIS          word1 tuberculosis word2
DIABETES              word3 diabetes word4
CHEMOTHERAPY          word5 chemotherapy word6
CARDIAC ABLATION      word7 cardiac ablation word8
ATRIOVENTRICULAR NODE word9 atrioventricular node word10
PULMONARY ABSCESS     word11 pulmonary abscess word12
6 rows selected.
SCOTT@orcl_11gR2> -- tokens that are indexed:
SCOTT@orcl_11gR2> select token_text from dr$class_phrase_idx$i
2 /
TOKEN_TEXT
ABLATION
ABSCESS
ATRIOVENTRICULAR
CARDIAC
CHEMOTHERAPY
DIABETES
NODE
PULMONARY
TUBERCULOSIS
9 rows selected.
SCOTT@orcl_11gR2> -- searches using text index:
SCOTT@orcl_11gR2> set autotrace on explain
SCOTT@orcl_11gR2> select phrase, document
2 from   classifications
3 where contains (phrase, 'tuberculosis') > 0
4 /
PHRASE                DOCUMENT
TUBERCULOSIS          word1 tuberculosis word2
1 row selected.
Execution Plan
Plan hash value: 2513347404
| Id | Operation                   | Name             | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |                  |     1 | 2046 |     4   (0)| 00:00:01 |
|   1 | TABLE ACCESS BY INDEX ROWID| CLASSIFICATIONS |     1 | 2046 |     4   (0)| 00:00:01 |
|* 2 |   DOMAIN INDEX              | CLASS_PHRASE_IDX |       |       |     4   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - access("CTXSYS"."CONTAINS"("PHRASE",'tuberculosis')>0)
Note
   - dynamic sampling used for this statement (level=2)
SCOTT@orcl_11gR2> select phrase, document
2 from   classifications
3 where contains (phrase, 'cardiac ablation') > 0
4 /
PHRASE                DOCUMENT
CARDIAC ABLATION      word7 cardiac ablation word8
1 row selected.
Execution Plan
Plan hash value: 2513347404
| Id | Operation                   | Name             | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |                  |     1 | 2046 |     4   (0)| 00:00:01 |
|   1 | TABLE ACCESS BY INDEX ROWID| CLASSIFICATIONS |     1 | 2046 |     4   (0)| 00:00:01 |
|* 2 |   DOMAIN INDEX              | CLASS_PHRASE_IDX |       |       |     4   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - access("CTXSYS"."CONTAINS"("PHRASE",'cardiac ablation')>0)
Note
   - dynamic sampling used for this statement (level=2)
SCOTT@orcl_11gR2> select phrase, document
2 from   classifications
3 where contains (phrase, '%ab%') > 0
4 /
PHRASE                DOCUMENT
DIABETES              word3 diabetes word4
CARDIAC ABLATION      word7 cardiac ablation word8
PULMONARY ABSCESS     word11 pulmonary abscess word12
3 rows selected.
Execution Plan
Plan hash value: 2513347404
| Id | Operation                   | Name             | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |                  |     1 | 2046 |     4   (0)| 00:00:01 |
|   1 | TABLE ACCESS BY INDEX ROWID| CLASSIFICATIONS |     1 | 2046 |     4   (0)| 00:00:01 |
|* 2 |   DOMAIN INDEX              | CLASS_PHRASE_IDX |       |       |     4   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - access("CTXSYS"."CONTAINS"("PHRASE",'%ab%')>0)
Note
   - dynamic sampling used for this statement (level=2)
SCOTT@orcl_11gR2> -- searches using non-text index:
SCOTT@orcl_11gR2> select phrase, document
2 from   classifications
3 where phrase = 'PULMONARY ABSCESS'
4 /
PHRASE                DOCUMENT
PULMONARY ABSCESS     word11 pulmonary abscess word12
1 row selected.
Execution Plan
Plan hash value: 4202264836
| Id | Operation                   | Name              | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |                   |     1 | 2034 |     1   (0)| 00:00:01 |
|   1 | TABLE ACCESS BY INDEX ROWID| CLASSIFICATIONS   |     1 | 2034 |     1   (0)| 00:00:01 |
|* 2 |   INDEX RANGE SCAN          | CLASS_PHRASE_IDX2 |     1 |       |     1   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - access("PHRASE"='PULMONARY ABSCESS')
Note
   - dynamic sampling used for this statement (level=2)
SCOTT@orcl_11gR2> select phrase, document
2 from   classifications
3 where phrase = 'PULMONARY ABSCESS'
4 /
PHRASE                DOCUMENT
PULMONARY ABSCESS     word11 pulmonary abscess word12
1 row selected.
Execution Plan
Plan hash value: 4202264836
| Id | Operation                   | Name              | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |                   |     1 | 2034 |     1   (0)| 00:00:01 |
|   1 | TABLE ACCESS BY INDEX ROWID| CLASSIFICATIONS   |     1 | 2034 |     1   (0)| 00:00:01 |
|* 2 |   INDEX RANGE SCAN          | CLASS_PHRASE_IDX2 |     1 |       |     1   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - access("PHRASE"='PULMONARY ABSCESS')
Note
   - dynamic sampling used for this statement (level=2)
SCOTT@orcl_11gR2> select phrase, document
2 from   classifications
3 where phrase like '%AB%'
4 /
PHRASE                DOCUMENT
CARDIAC ABLATION      word7 cardiac ablation word8
DIABETES              word3 diabetes word4
PULMONARY ABSCESS     word11 pulmonary abscess word12
3 rows selected.
Execution Plan
Plan hash value: 723026238
| Id | Operation                   | Name              | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |                   |     3 | 6102 |     0   (0)| 00:00:01 |
|   1 | TABLE ACCESS BY INDEX ROWID| CLASSIFICATIONS   |     3 | 6102 |     0   (0)| 00:00:01 |
|* 2 |   INDEX FULL SCAN           | CLASS_PHRASE_IDX2 |     1 |       |     0   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - filter("PHRASE" IS NOT NULL AND "PHRASE" LIKE '%AB%')
Note
   - dynamic sampling used for this statement (level=2)
SCOTT@orcl_11gR2>

Improved Keyword Searching Needed

One frustration I have with Lightroom is with the difficulty it has with dealing with spaces in keywords when trying to do searches. For example, if I have images with the keyword 'University of Michigan' (no primes around the keyword), I'm unable to find images with just the keyword 'University of Michigan'. Instead, I'll get things with University, Michigan, etc. The only solution at the present time seems to be to use underscores or something equivalent in multi-word keywords, but this seems silly and ugly. I should either be able to search for a quoted keyword and have the search complete, or Lightroom should use the comma as a keyword separator in the Find dialogs in the same manner as it is used when adding keywords. My main reason for using Lightroom was to be able to use its database-like features for keywording and finding images based on keywords (and other attributes). Dealing with multi-word keywords should be a simple fix and it would be greatly appreciated.
Thanks!
James Howe

I'm trying to build a smart collection so I'm limited to searching for 'Contains', 'Contains Words', 'Contains All', 'Start With', 'Ends With'. No matter which option I select, I end up getting images which don't match. I only want to see images which have the specific keyword 'University of Michigan' (a multi-word keyword without the primes). However, the search treats this as a search for images with the individual keywords of 'University' 'of' and 'Michigan' and therefore finds images which not only have the keyword 'University of Michigan', but also any image which has keywords of 'University', 'of' and 'Michigan' (the results vary depending on which search criteria I pick). The important thing is that there is no query that I can come up with which *only* finds those images with the specific, single keyword of 'University of Michigan'. If I change the keyword to be 'University_of_Michigan', I can find that quite easily. The problem is with the inconsistent way that Lightroom treats spaces in keywords. If I go to the keywords entry field, I can enter something like:
Foo, Bar, Keyword with Spaces, Baz
and this will add four keywords to the image. However, when I try to find or build a smart collection based on keyword values, if I use the keyword
'Keyword with Spaces', the search algorithm treats this as three keywords, not a single keyword. Depending on the type of search selected (Contains, Contains Words, Contains All), the search engine will look for images with the keywords 'Keyword' AND 'with' AND 'Spaces' or 'Keywords' OR 'with' OR 'Spaces', etc. I believe that Contains Words comes closest, but in addition to finding images with the keyword 'Keyword with Spaces', it will also find images with the keyword 'Keywords with Spaces and other stuff'. If I try to quote the phrase, the search engine won't find it because the keyword wasn't defined with quotes (or primes). All that is needed to fix this is to provide some mechanism which allows the user to indicate that a collection of words should be treated as a single keyword, not a group of keywords.

Keyword filtering problem in Smart Collections

All,
I’m trying to do create a Smart Collection based on Keywords, unfortunately I can’t seem to get it to work. The criteria is that the collection must contain the keyword phrase “Publish to Flicker & Facebook” but not “Published Image”.
I’ve tried variations of the following:
Match all of the fallowing rules
Keyword doesn’t contain = Published Image
Keyword contains = Publish to Flickr & Facebook
This should give me all the images that I want to Publish to the services but have not been published as of yet. If I do either of these separately they work but not together as a filter. FYI I’m using the Beta 3 version. - CES

That didn't work... even renamed to nonsence words still dosent work

Keywords in Adobe Bridge CC and Lightroom 5 not working

I recently upgraded to Adobe CC. We use Bridge and Lightroom regularly to tether while shooting, batch process and search for images. I have noticed that files I have keyworded in Bridge will randomly lose the keywords attributed with them. When we shoot tethered to lightroom we will keyword as we shoot but those keywords do not show up in Bridge. This is a huge issue for us. We have thousands of images and plan on using bridge as a tool for our designers and marketing team to search seamlessly for photos based on keywords. If they are randomly disappearing or not carrying through from Lightroom to Bridge it becomes completely unmanageable and completely out of our hands. And a completely useless tool. Any suggestions?

Keyword list in Adobe system, bridge and lightroom, need standardizing. That's what I feel.

How to get Port Classifications for Logical Switch via scvmm console

Hello All,
We are creating some automation tool based on SCVMM console scripts.
In order to create vm we need to assign port classification based on Virtual Switch and then logical switch. The problem is that I could not find any way to retrieve port classifications assigned to logical switch (Virtual Port tab in SCVMMs logical
switch properties). There is Get-SCLogicalSwitch that return just name, description and some properties. There is also Get-SCUplinkPortProfileSet which returns Uplink tab data of SCVMMs logical switch properties. But no api to retrieve port classifications
for switch. Get-SCPortClassification also doesn't return switch assignments. Any ideas how to get this data?

I think you are looking for
Get-SCVirtualNetworkAdapterPortProfileSet.
This posting is provided "AS IS" with no warranties, and confers no rights. User assumes all risks.

CS4 Bridge Keywords - Delete when Bridge is close.

Greetings
I have gone through and applied keywords to hundreds of images, along with a respective Hierarchy KeyWords as well. As long as Bridge remains open and my computer is on I have access to sort, and find files based on Keywords.
"Close Bridge" : and all Keywords are missing from Keyword pane, excluding the few that are present with Bridge as examples that are always there.
I can, re-select all the images, and the keywords will load again into the Keyword pane, but that takes about hour to manage.
It is like some preference file is not being read, or the Keywords are not saved to the Keywords pane. I have loaded again this App from Photoshop CS4.
One person suggested I go into Preferences and un-click "Read Hierarchy keywords" then open all my image folders and again click on the images to reload the keywords into the Keyword pane again. Then go back to preferences and re-click Read Hierarchy Keywords again, and that everything should be in order. I have done this twice, once with starting Bridge again before re-clicking RHK button again.
Still after two years of having Bridge CS4, I can not used this function to search for images.
And yes, I do select a folder or mounting disk for Bridge to look at for finding the Keywords.
After all this time, has anyone solve the riddle of trying to make this work. I can not update my MAC. I have a PowerPC, Quick Silver, 2002 G4, dual 1000. So 10.4.11 is the last operating system I believe I can use.
Thanks for you help.
Have a good day.

Hello Curt
Thank you for your reply.
This Keyword system was working earlier, that is I could shut off the computer and come back to open Bridge and the keywords were there.
As you suggested the keyword did come up in Italics as I began working with them a few days ago, because someone had responded to an old Thread of mine from this site from almost a year ago. Initially there were coming up as persistent in normal font face.
I followed your suggestion with a smaller batch of Keywords, then shut down the computer and, and they were present when the program came back up.
So Thanks, that works.
This will get me back on the right track again with Keywords.
I have no idea what iniciated a change of from persistent keywords to loosing them all in the past. But at least cognitively, I now have a clear process for getting them to show up again on start.
Thanks again for your reply.

Classification based on "keyword phrase"?

Similar Messages

Maybe you are looking for