============================================== Version 3.3u2: 31 May, 2019 ============================================== Fixed the algorithm which splits texts into display lines in the Coding window, so that Chinese text lines don't go off the page (Thanks, Tenglong). ============================================== Version 3.3u: 20 April, 2019 ============================================== * Fix "Apply Gazette" so that system does not look for matches within multiword segments. (Thanks, Morgane) ============================================== Version 3.3t: 16 April, 2019 ============================================== * Change of Installer Maker to make (hopefully) the installed program actually work on Windows. * Further small fixes to Tagger which failed in some contexts. ============================================== Version 3.3s: 10 April, 2019 ============================================== LAUNCH * On some Windows systems, the system would fail to lauch due to version incompatibilities with a file "stringprep". This is (hopefully) resolved. APPLY GAZETTE Yet more small changes to version 3.3r to fix problems with Apply Gazette. i) Where the tagging of one file fails, it is reported, but the application continues with other files. ii) When tagging does fail, the reason for failing is reported. iii) Sometimes the reason for failing to tag a file was a programming error. Now fixed. AUTOMATIC THEME ANALYSIS The automatic Theme layer was inoperable in version 3.3r. Should now be fixed. ============================================== Version 3.3r3: 1 April, 2019 ============================================== Small changes to version 3.3r to fix problems with Apply Gazette. ============================================== Version 3.3r: 25 March, 2019 ============================================== Revised import metadata option under the "Project" menu. This now requires the metadata file to be an Excel file (xls or xlsx). Revised the Save Glossary and Apply Glossary commands (Misc menu). These now work with Excel files. This avoids problems from the previous version which used tab-delimited text, and the user needed to ensure the text file was in utf-8 encoding. ============================================== Version 3.3p: 12 February, 2019 ============================================== * For those depending on TreeTagger to tag files, the parameter files for different languages moved on the web. UAMCT now finds them again. Would not affect you if you already have used the language you are interested in with UAMCT. ============================================== Version 3.3p: December, 2018 ============================================== * Some fixes for making a distributable install properly on Macs running Maverick. May not have worked. ============================================== Version 3.3o: 5 October, 2018 ============================================== * Fixed a bug in CQL search affecting search for segment within segment. * Added in "Save Gazette" option under the Misc menu: saves a tab-delimited file representing the most common coding of each text strings on a layer. The gazette can then be applied to new files on that layer as a form of autocode. ============================================== Version 3.3n: 12 March, 2018 ============================================== * Documents in encoding "936" were not dealt with by Stanford tagger. This has been fixed (Thanks Jianwei) * Autocode of documents very slow due to displaying whole text in window. Now quicker as only first 100 chars shown. ============================================== Version 3.3m: 12 November, 2017 ============================================== BUG Fixes * On some versions of Windows, download of the Stanford Tagger and Parser failed due to some problem in the installer. Should now be fixed. * Layers specified as "token" segmentation no longer auto-segmented. Now fixed. To apply the segmentation to existing files, open the file for annotation at the token layer, and select "Resegment Document" from the "Codings" menu. (Thanks Mario) ============================================== Version 3.3l: 6 October, 2017 ============================================== * With layers defined as "error" layers (when the layer is created), you can now search for the segments with a particular correction, e.g., CQL: Or with the widget based approach, click on the + and select "with_field" * Exporting annotations in XML format: the "correction" field of error layers is now exported as well. ============================================== Version 3.3k: 17 July, 2017 ============================================== * Autocode now shows 3 tokens of context on each side of the hits * SEARCH: Search for "X within! Y" (or "X not_within Y") now works. (was broken) ============================================== Version 3.3h: 17 November, 2016 ============================================== Minor fixes in the Stats and Explore pane Statistics: - Fixed the "Lematised Wordings" option. Explore: - where files yet untagged, the program reported a problem on each one. Now tagged automatically, as with Search and Stats. - Propensity: previously, any token with ============================================== Version 3.3g: 17 October, 2016 ============================================== BUG FIX RELEASE SEARCH, STATS * Ignored segments should not be found in search or statistics. Now fixed (Thanks Amanda) STATS: * Fixed bug using Feature stats: per 1000 words. Should now work. ============================================== Version 3.3f: 8 April, 2016 ============================================== BUG FIX RELEASE * Fixed bug in Statistics where comparing sets and a set used "in_segment" (thanks Mercedes) * Fixed problem with tabs in comments of segments (Thanks Miguel Angel) ============================================== Version 3.3e: 8 April, 2016 ============================================== BUG FIX RELEASE * Fixed problem in search where searching for unit in another unit (Thanks Andrea) * Explore pane was nonfunctional, now (mostly) working (at least 'words' and 'features'). * Adding A Stanford POS tag layer, specified as extended, the program would wrongly give a syntactic analysis layer rather than the POS layer as requested. Now fixed. * Added some words to English closed-class list: very, through, above, behind, below, up, via, nary, thy, &. * Statistics: when selecting "Describe each file", and the unit of interest was a document-level feature, the results wrongly included files which did not match that document feature. Now fixed. * In subjective stats, when the subjectivity was perfectly balanced (a 0 score), a hyphen was shown, indicating "not calculated" by mistake. ============================================== Version 3.3d: 8 April, 2016 ============================================== * Color coding of a text (accessed via clicking on the filename of a file in the Files pane) was broken in the last few versions. Now fixed. ============================================== Version 3.3c: 6 April, 2016 ============================================== BUG FIXES * After running a Statistical study, the diagnostic window stayed open. Now shuts when complete. * The automatic POS layers were not functioning in the last release, all the most delicate tags were correct, but the universal features (noun, etc.) were not assigned. Now fixed. If you had this problem, you may need to delete your POS layer and add it again to get things right. * The total token count given in Lexical stats was sometimes out by 1, as the last character in a text was dropped by mistake. Now fixed. A file with text: "The cat sat." should have four tokens. ============================================== Version 3.3b: 10 March, 2016 ============================================== * In General Stats, the item labelled 'Words per Segment' should have read 'Tokens per segment', as the count included punctuation symbols. Now, the software provides both 'Words per Segmen' AND 'Tokens per segment'. * SEARCH using WITHIN!, MATCHING! and CONTAINING! failed to filter. Now fixed. * Added a Modality layer to recognise Modality tokens in English. * Improved Mood and Extended layer for recognition of it-clefts, extraposition, etc. Still not perfect (over-recognises) * Changed the Automatic Theme layer to recognise elliptical themes (I like Mary but <> her cooking) and empty themes (THERE is a cow in the garden; IT is John who is late) ============================================== Version 3.3: 10 November, 2015 ============================================== SEARCH: Major overhaul of the search functionality, which should better integrate with Stanford Tagger and TreeTagger when tagging is needed. TAGGING: * Better placement of the tagger tags (nns, etc.) into the wordclass hierarchy (scheme) provided with POS layers, and better handling of tags for languages other than English being integrated into the scheme. * For treetagged files, [, ], ", ', and other qyuote marks were not recognised as separate from the word they appeared next to. Now fixed. SCHEME * Fixed a rare problem whereby the menu attached to features and systems would not pop up (Thanks Carolina). ============================================== Version 3.2j: 11 November, 2015 ============================================== Scheme: * Fixed bug where in some cases, clicking on a feature or system would not produce the menu (Thanks, Carolina) Explore: * Added Log Likelihood statistic to Keyness table Statistics: * fixed bug in saving Wordings to file. Compare Codings: * Where segments differ by only white space (space, tab, newline), they were considered different segments. Now considered the same (Thanks, David) Help Menu: * On windows sometimes Manual and Tutorial would not show, now fixed. ============================================== Version 3.2i: 30 September, 2015 ============================================== * Fixed bug whereby the Extended Mood layer failed (Thanks Leigh) ============================================== Version 3.2h: 15 July, 2015 ============================================== * Fixed bug when using TreeTagger to POS-tag files, would not work (since 3.2f or so) ============================================== Version 3.2g: 15 July, 2015 ============================================== * Fixed bug in Stats, whereby clicking on the feature count numbers opened Search for that feature, but broke. ============================================== Version 3.2f: July, 2015 ============================================== * Totally revised the explore pane. So far only word and feature are back in, with frequencies or keyness. * Fixed a bug in Search and Stats such that any segment which extended to the last char of a file were ignored. (except for document layers) ============================================== Version 3.1.17: February, 2015 ============================================== Various bug fixes; * Text Styling was broken, should be fixed. * Stats/Compare Files/General was broken, now fixed. ============================================== Version 3.1.16: December, 2014 ============================================== * Some bug fixes in the search code. - when searching for sequence of segment specifications, sometimes would match when shouldn't, with punctuation marks between the matched segments. CODER WINDOW * On Mac under Yosemite, the menus in the Coder frame didn't display properly. Now fixed. PARSER: SFL-MOOD * Presence of a 'Predet role did result in feature 'predetermined-group' being added to the phrase. Role is now 'PreDet' and the feature is added. * Changed analysis so that 'all of' in 'all of my friends' and similar is recognised as a PreDeterminer. * Fixed handling of 'be' as main verb (copula in Stanford) * Fixed Nom group conjunctions of more than 2 elements. ============================================== Version 3.1.15: September 29, 2014 ============================================== * Overhall of the interface to TreeTagger to ensure tagging works ok. This will resolve many problems you might have had. Now working for: english, french, german, dutch, italian, spanish, russian, bulgarian, finnish, estonian", polish, slovak, latin, mongolian, swahili, galician. ============================================== Version 3.1.14: September 15, 2014 ============================================== EXTENSIONS * Coloring segments in annotation: to color-code the segments in annotation windows, open the scheme, and click on a feature and select 'Add Realisations' and add a line like: color red From then on, segments coded with that feature will be displayed with that colour (may need to close and reopen the coding window for changes to take effect, or in some other way cause the window to be redrawn). * Extend Corpus: added a 'Paste' button when the Paste from Clipboard option is selected. Previously, only ctl-v (0r cmd-v on Mac) was possible. * Previously, when adding a layer and selecting the built-in scheme for Appraisal, only a combined Attititude, Graduation and Engagement scheme was available. Now 4 schemes are available, offering schemes also with just Attitude, Graduation or Engagment by themselves. BUG FIXES * Import Metadata broke if the imported spreadsheet had accented characters. Now fixed. * Import Metadata: if column headers, imported as field names, included a space char, the program would break. these are now automatically replaced by '_' ADD LAYER * The built-in scheme for 'Clause grammar' was missing, now provided. SEARCH * The stop button during Statistical analyses should now work * Where search for query1 IN query2 and the second query has a syntax problem, the search would just break. Now fixed. STATS * Fixed problem displaying bar chart mode, per 1000 words, where a category had zero words (diagram just didn't draw) ============================================== Version 3.1.13: August 31, 2014 ============================================== NEW FEATURES * Export Layer: Previously the feature to export a layer was available from the Project menu (although broken). This saved each file separately. A new option was added to save all codings of a layer in a single file, tab delimited. This can then be used in Excel or similar to compile statistics. * Search: Added "Containing CQL Query" to the options available in the menu-driven search interface. ANNOTATION * restored the "Delete" button to the toolbar, as having to get it from a menu was annoying. BUG FIXES: - ANNOTATION: * When in Table Annotation mode, double clicking will bring up a list of features to choose from for the segment. This list is now sorted to match the order from top to bottom in the scheme. * Auto-advance to the next/previous incomplete (<<, >>) was broken. Now fixed. * Sometimes the selected segment was off the page, not visible. Should be fixed. - EXPORT ANNOTATIONS * Selecting "ExportLayer" from the Project menu and then indicating "Tab Delimited" format was broken. Now fixed. - TREETAGGER Tagging * On first use of TreeTagger, the software now tries to run the file you identified as the TreeTagger executable, and will complain if it cannot run. SEARCH * Fixed small bug when searching for feature1 and not feature2. AUTOCODE * Fixed problem with delete segments ============================================== Version 3.1.12: August 18, 2014 ============================================== TAGGING * Fixed some bugs over the prior version that interfered with creating tagged data for new files. * Added option in the Options pane (Taggging Options) to change the location of the TreeTagger executable. SEARCH * Revision of search code for segments, now 4 times faster. * Fixed bug in saving Search results, where the saved table became misaligned if search results included html tags. * Fixed bug in display of search results where sometimes hundreds of results shown in each page (should be just a page full, with no scroll). STATS * Added "Wordings" option in Statistics, which lets you track how the use of different words or phrases vary over multiple sets. * Deleted the "Compare Two sets option". The same option can be achieved by "Compare Multiple" with 2 chosen. * Added a "Structure" annotation layer which provides a better annotation interface for many applications. (not yet documented, but when creating a layer, choose "Manual Annotation", "Segments within a document" then "Structure".) ============================================== Version 3.1.11: July 4, 2014 ============================================== CQL * use of [] to match a token of any kind was broken, now fixed. ============================================== Version 3.1.10: June 28, 2014 ============================================== EXPORT * Modified the Export Layer function (from the Project menu) in various ways: - now you are not prompted for a save destination, saves the exported files in the Exported folder of your project. - added a "Save in tab-delimited format" which is easy to read for humans and for programs. CQL SEARCH * When searching for segments with a feature using the shortcut form: , where the feature name is available in to layers, previously only the first was searched, which in some cases led to confusion (the system seemed to not find valid matches). Now, the tool will warn you and tell you to specify the layer. ============================================== Version 3.1.9: June 26, 2014 ============================================== SEARCH * CQL searches using term 'not_within' (or within!) now properly eliminate those hits which were within the specified unit. * CQL search: the interface now recognises illegal token attributes, (e.g., [lama="dog"] where 'lemma' was intended), reports it and stops (instead of doing a long search with null results) * Clicking on a presented segment would open the coding window for that segment, but then the coding window would dissapear. Now fixed, stays on top. STATS * In feature stats, you can click on the count of each feature, and the interface will switch to the Search interface showing the instances which the count includes. However, if you had the Search interface switched to "Menu-based search", then the Search interface wouldn't update. Now fixed. * Fixed various small bugs in Stats introduced in 3.1.8, e.g., general stats with a token layer would lose all token segments, per 1000 would give bugs if a file had no hits in it, etc. ANNOTATION: * Fixed bug where the lines indicating a segment would continue to show after the segment was deleted. LAYERS * Added a "Structure" layer for manual annotation which appears in the coding window much like the automatically parsed layers. Select text to create segments, but displayed with boxes under the text. Double click on a box to choose the feature. Drag the side of a box to change its size. ============================================== Version 3.1.8: June 17, 2014 ============================================== MAJOR UPDATE OF STATISTICS TO HANDLE LARGE CORPORA (Thanks Daniel McDonald) - reworked the Stats code to avoid high memory usage when dealing with large corpora (statistics are derived accumulatively from each corpus file in turn, rather than finding all matching segments in the corpus and then deriving stats). - per 1000 word stats: this will now use the tagged data for each file (using either Stanford tagger or TreeTagger, or if neither of these available, using my own internal tokeniser). Should make per 1000 calculations much quicker, once all files are tagged. - General Stats: this will now give some stats in proportion to WORDS (alphanumeric for western languages) not for TOKENS (which includes punctuation). As a result, some figures will now be a slightly different, e.g., lexical density a little higher, syllable length higher, words per segment lower. SEARCH: searching on fields and corpus/subcorpus was not connected. Now working. ANNOTATION: Added a new mode of annotating texts, allowing creating new segments while in Structure view, (by selecting the text), and coding directly by clicking on the segment label in the structure display. AUTOCODE: specifying feature "A and not B": the "not" would be lost somehow, so searched for "A and B"?. Now fixed. TAGGING: sometimes the program would try to ensure the tagger was installed even if the whole corpus was already tagged. Now the program should only try to install the tagger if it is actually needed at the moment. ============================================== Version 3.1.7: May 27, 2014 ============================================== * General checkover to remove remaining major problems in Statistics and Explore. - Statistics should be generally functional except for save as PDF - Explore should be ok for Exploring single text (not sure of heat view). Explore sub corpus not full checked yet. - Explore in the special Appraisal mode not yet checked. ============================================== Version 3.1.6: May 1, 2014 ============================================== *IMPROVEMENTS* STATISTICS - Save of General Stats to file did not save two of the values displayed on the screen, min/max segment length. Now fixed. - Network view of feature stats always displayed the entire network, even if the unit of interest was somewhere deeper in the network, meaning that much of a network had no results displayed. The code was modified to only present the part of the scheme that is relevant. - also, glosses are not displayed by default. EXPLORE * Word counts in table format now also shows relative frequency (number of occurrences of the word divided by the total number of words in the text. * BUG FIXES* AUTOCODE * This feature, broken since the start of 3.1, is almost functional again. MENUS * "View Release Notes" from the Help menu was broken, now fixed. ============================================== Version 3.1.5: April 5, 2014 ============================================== NEW FEATURES Coding Window: * For structural layers (Stanford parser or any of the automatic syntax layers), you can now move between a table-like view of the structure and a tree structure. * In both view modes, different editing options are available. Tree view: - drag the red dot above a segment and drop it on the box of another constituent to change its parent. - Double click on a box to see a list of actions on that tree. Table View: - Drag on the borders of a box to expand/contract a segment. - Click on a role name to edit the role name. BUG FIXES * Search: - fixed bug in searching for something "within" a document, e.g., within . This broke before, now fixed (Thanks Daniel) - use of "Feature and not Feature" in menu-driven search has been broken in earlier versions of 3.1. Now Fixed. * Coding: when a coding window is open, and you open the scheme editor and delete a feature that had already been selected, it was not deleted from the list of selected features. Now fixed. DOCUMENTATION * Added documentation of Fields, Import Metadada, and extended documentation of CQL Search. ============================================== Version 3.1.4: March 8, 2014 ============================================== New Features: * Key combination "Control-w" (Cmd-w on Mac) will close certain windows (Coding windows, colour coding windows, Popup Help windows). * Changes to Coding window: - if there are multiple choices that can be selected from at this point, you can change to one of the others by clicking on the system name above the choice and choosing one of the others. - Where the annotation layer involves each segment having a "role" label, then this can be edited in the space on the right of the pane ("Role"). You can hit the Return key to display your change. * Search Window: Added Summarise as a search display option, to show how frequent each particular result is. ============================================== Version 3.1.3: March 3, 2014 ============================================== BUG FIX RELEASE * Several fixes to bugs in general Search code. * Statistics now working again (Autocode still broken) * Explore working for lexical searches, not for feature-based searches. ============================================== Version 3.1.2: February 12, 2014 ============================================== BUG FIX RELEASE SEARCH: * lemmas with non-ascii characters would not match in search due to unicode problem. Now fixed. * searches of segments CONTAINING something would return false matches in some cases. Now fixed. * Clicking on the text of a segment match should open the coder window with the segment highlighted. For POS windows, selected segment was not highlighted. * Where two words are not separated (e.g., "he's" or "nest" in French), where the first word is displayed as pre-context for a hit, the last character was not displayed. AUTOMATIC ANALYSIS * Recognition of relative clauses was going wrong within the SFL-Mood analysis. ============================================== Version 3.1.1: February 12, 2014 ============================================== GENERAL BUG FIX RELEASE * On MacOSX, some windows would not "lift" to visibility, but instead stayed behind the main window. Now fixed. * "Export Anotations" from a Coding window saved to a strange file location. Fied. * Mood parsing: an prep-phrase adjunct in a clause with a ditransitive verb would always be recognised as the indirect object. Now limited to "to" and "for" prepositions. This had a roll-on effect on Theme and Transitivity parsing for this element. * Opening of project: for projects with large numbers of files (> 100), the opening was someone slow. This has been improved. * Import UAMCT2 projects: in some cases, segments would be imported as ignored, now fixed. * "Save Student Feedback" from within an Error coding window didn't display properly if the comment or error correction has a ' or ". Now should be fixed (Thanks Fiorella). ============================================== Version 3.1: February 2, 2014 ============================================== * Radical overhaul of features to support needs of Open University and the French Department at the UAM. * All text files will be automatically tokenised, and POS-tagged where a tagger exists. The default tagger will be the Stanford Tagger (works for English, French, German, Chinese and Arabic). If you install TreeTagger (Helmut Shmid), and tell UAMCT to use it, then you can tag using this instead, covering English, French, German, Italian, Spanish, Bulgarian, Dutch, Galician, Estonian, Swahili, Slovak, Latin, Mongolian an Finnish. Other languages will be tokenised but not tagged. * Added a special layer for POS-tagging, which displays in a nice way, and allows you to edit the POS tagging where the automatic tagger went wrong. This layer allows you to treat pos-tags the same as any other annotation layer (e.g., search, statistics and explore with parts of speech). * Stanford parser for French and German at least should work, still need to test/debug other languages. * SEARCH: Search now supports CQL syntax (Christ and Schulze, 1994). (see: http://www.wagsoft.com/CorpusTool/Documentation/CQLSearch.html) which allows you to search for a wider range of things that was possible before. The old menu-based search is still available. MINOR CHANGES * Added "Change font size" to the "other action" menu in Coding windows (to facilitate quicker changing of display size) * On Macosx, Snow Lion, clicking on a .ct3 file would not open the file but show the Opening window of UAMCT. Fixed. ============================================== Version 3.0 Beta 5: September 30, 2013 ============================================== * Using concordance with files which had an incorrect encoding specified would lead to these files being retagged on every search. Now fixed. * Text Styling of text (click on the filename in the Files pane) now allows you to specify text background colour. * Scheme editing: fixed a bug which could happen when pasting a feature name into the Scheme, where the pasted text included a newline char or similar. ============================================== Version 3.0 Beta 4: August 30, 2013 ============================================== * Added Automatic POS layer, and a nice coding window for it. * Fixed problem in calculating Subjectivity strength. (polarity neutral items were treated as neutral in strength, but they may not be). * STATS: saving a feature study, percent and N were reversed (thanks Thu). ============================================== Version 3.0 Beta 3: June 30, 2013 ============================================== * AUTOMATIC ANALYSIS: - Fixed occassional failure where parser could not parse a sentence, the whole file couldn't be Theme or Mood analysed (thanks Ryan) * Added automatic Transitivity analysis (still quite a lot of problems) * Automatic Theme analysis: - Switched function label from "Ideational" to "Topical". * Automatic Mood analysis: infinitive clauses were accidentally given some finite clause features. Now fixed. (but you will need to select "Re-segment document" from the "Other Actions" menu on the coding pane, to cause each sentence to be re-analysed) * Added "help" pages for the "Files Pane" and the "Layers" pane, so the "help buttons on these windows now works. * Replaced the code used to display Help text, so bulleted items now display as bulleted, not as numbered items. * Storing of field data (e.g., the fields of subcorpus and filename for each document-layer) was being dropped. Now fixed. * CONCORDANCE SEARCH: for some searches, the "context" setting was ignored (how many words around the match to show. Now fixed. This affects those searches which contain no actual words, e.g. "@vb* @vbn". * File Pane: colour coding showing file annotation completeness has been reversed. Buttons are most salient when some coding is left to do on the layer/file. * In general, white text on light background eliminated. * Coding frame: sometimes text would be off the Screen to the right by one word, now fixed. (Thanks Ryan) ============================================== Version 3.0 Beta 2: June 17, 2013 ============================================== DOCUMENTATION * A Tutorial guide to the basics of UAMCT3 is available from the Help menu in the menu bar. FILE PANE * Vertical Scroll of File Pane didn't work. Now fixed (thanks Andrea) * Color coding of file buttons to reflect degree of coding on the file/layer should now work properly. CREATE NEW LAYER * Specifying use of one of the built-in error Schemes did not preselect the layer as an"error" layer (which provides a space for the coder to type the correction of each error) CONCORDANCER * Added a Help page for Concordance searching, also listing the POS categories available for English. * Added a Help button on the Concordancer page will will show the Help page. CODING FRAME * Feature in the "Selected" box are now displayed sorted in terms of delicacy (thanks Anke) AUTOMATIC ANNOTATION * Automatic Theme annotation now added several features to each theme-group: has-textual-theme vs. no-textual-theme, has-interpersonal-theme vs. no-interpersonal-theme and unmarked-ideat-theme vs commplem-ideat-theme or adjunct-ideat-theme. With these, it is possible to do some quick comparison of textual patterns between sub corpora. Select "theme-group" as the unit of interest. (if you already have a theme layer, you will need to delete it and add a new one to make this work). ============================================== Version 3.0 Beta 1: June 15, 2013 ============================================== Changes from version 2.8 are many. Major areas of change are summarised below: * Launch: Opening the project window now much quicker, especially where 100s of files involved. * PROject Window is gone, replaced by separate "Files" and "Layers" pane, each with more space. * Search, Stats: should be up to 10 times quicker now, This is due to a change to the way annotations are stored (see bleow) * DATA FORMATS: UAMCT no longer uses XMl to store files. Binary storage format used. Option to save the project as UAMCT 2.8 format is supported to use XMl format. This new format also avoids a bug whereby search returns wrong segmnts due to the index files getting out of sync. AUTOMATIC ANNOTATION * Added layers for automatic annotation of Theme SFL Transitivity. ADD NEW LAYER * Revised the Add Layer interface. ADD FILES * Added an option to add a file via Paste from Clipboard, a quick way to get a text into UAMCT. TEXT EDITING * Added an option "Edit Text" in the "Other Actions" menu for a file on the Project window. You can edit the text of a file, and most of your codings of the file will be adjusted accordingly. Doesn't allow cut/paste of text though, just inserting or deleting via typing. SEARCH: * Added a concordance search option, which allows searching on parts of speech, using POS-tagging of your files. Lightning quick. Your files are POS-tagged courtesy of Stanford PSO Tagger. * A new display for concordance search, presenting the KWIC data as a graph. AUTOCODE * Added an option to delete all segments matching a certain pattern, e.g., delete all segments with a specified feature. STATISTICS * TSTATS no longer shown for feature usage comparisons, as it is not appropriate for this type of data. EXPLORE * Special Attitude Interface: Where you use an Attitude layer which includes features: affect, appreciation and judgement, this will trigger the inclusion of a special interface for exploring Attitude patterns in a text. NOVICE MODE * There is a "Novice mode" which reduces the windows provided with CorpusTool, hiding the Project, Autocode and Options windows. Where students are using the software to explore a corpus provided by a teacher, this reduces the complexity of the interface they need to learn. If you provide select Novice mode before sending students a project, the project will open in Novice mode on their machines.