Typecraft v2.5
Jump to: navigation, search

Help:How to annotate in TypeCraft - a practical guide

Revision as of 16:48, 21 September 2014 by Dorothee Beermann (Talk | contribs)

TypeCraft 2.0 Editor

Since the fall of 2014, TypeCraft offers a new editor which allows a better integration of text-and sentence-level annotations. While text-level annotation allows you to annotate for discourse propterties, here called Senses, the sentence-level annotation allows for Valence annotations on the global-level and for lexical and functional annotation of words and morphemes.

Consider the screenshot of the TypeCraft 2.0 editor (TC2-editor) below

Figure1 Editor Views illusting different annotation levels

In the background of screenshot to the left, showing the TC2-editor, we see the editor's text area which contains green, not yet instantiated, phrases and blue, instantiated phrases.

To the right of the screenshot, we see an input matrix which is used to enter Metadata information about the text.

In the forground to the left, the Sense annotation viewer is shown. It visualised Sense-tags though colour coding. By pointing at words also Sense-tags can be made visible.

The Interlinear Glossed Text (IGT) editor in the centre of the screenshot shows the different levels of annotation, either in the form of a string of annotations (for valence) or directly in form of a specific tier. Notice that the valence annotation is shown above the annotation table. It is called construction description, and the tags are referred to as "global tags".

We will now discuss how you can use the TC2 editor to add different layers of annotations to a text. Also the TypeCraft 1.0 editor is still available for use, and we will maintain it until summer 2015.

TypeCraft 2.0 search and export of data are discussed in seperate sections.

Using the TC2 editor for texts that have been created with the TC1 editor

When your open a text that contains or consists of material annotated in the TC1 editor TypeCraft will pop-up a window and ask:

Your text has been created in the old editor. TypeCraft recommends the new editor and would  like to convert this text before opening it in the new editor. 
Note that the conversion cannot be reversed. If you would like to know more about the new editor before you decide by clicking on YES or NO below, 
please go to the TypeCraft Help page (and a link to the present page will follow)

When pressing YES the text will open in the TC2-editor, otherwise the text will open in the old editor. We will maintain the old editor until summer 2015.

Create a new text

You are logged in. There are no texts in your home domain called *My texts* because you have not yet started to annotate.

This is how you create a new text and split it into phrases that you can annotate.

Step 1 - open Text Editor in New Tab

To start, click on *New text* which you find in the upper navigation box to the left of your browser window. Right click it, and choose the second option from the drop-down window saying *Open Link in New Tab*. Having the TypeCraft wiki and the TypeCraft Editor open in two tabs will allow you to go more easily back and forth between these two. You can for example keep the Help Page open while you work with your TypeCraft Editor. Your normally find your tabs on the upper part of your browser window.

Step 2 - enter text into the text field

Go to the Text Editor window either by choosing *New text* or *My texts* from your navigation bar. The Text Editor has on the left side a text field. You use this field to either to import text from a file by copy & paste, or you manually enter sentences.

NOTE - any collection of phrases is a text.
Under Text we understand either real text, or part of real text, for example from a newspaper or a book etc., but also ::a collection of phrases, either sentences or smaller constituents, such as noun phrases which have been collected to illustrate a certain linguistic phenomenon.

Step 3 - sentence split

After you have entered your text you click on *Create phrases*. A dialog box appears. If you have not marked parts of the text (which you normally only do if you want to add more text to an already existing text), TypeCraft will say: Nothing selected. Should TypeCraft use the whole text instead? Answer: yes.

(Step 4 optional) - repeat sentence split

If you are unhappy with the way TypeCraft has split your sentences, clean out your text by for example deleting spaces between paragraphs, or by inserting periods. You can repeat this step until you are satisfied with the sentence break up.

How to annotate individual tokens

Step 1 - Initializing your annotation table

Click on one of the individual blue phrases on the right-hand side of your Text Editor window. A small window pops-up, saying:

TypeCraft wants to know:
This phrase has no words yet.

You can initialise words and morphemes automatically. Separate words by spaces (" ") and morphemes by hyphens ("-").

If you click cancel you can insert words and morphemes manually.

It seems that we have added this window to confuse you ;); what we really mean is this:

You have 3 different option how to use the input mask that contains now the sentence that you want to initialize:

  • if you want the phrase in the input mask inserted into a table without any further segmantation, click OK.
  • if you want the text in the input mask inserted into the table and you in addition want to split some of the words into smaller segments,, you can do that by inserting hyphens "-" or spaces " " in the phrase. Then you click OK. Do not be afraid of possible mistakes you might make when inserting hyphens at this point. You can always change what you do later.
  • If you don't want to start off with the material in the input mask, but rather wish to fill all material into the table manually click cancel. A one-column skeleton of a table appears, and here you can fill in text in the top line, one word per column. You create a new column by clicking in an existing column: you get a menu where you left-click, and get the options 'New word before', 'New word after' or 'Delete word'. By clicking on either of the former options a new empty column appears, where a word can be written in. On the second line in the table, morphological units are written in in a similar manner - the menu now offers 'New morpheme before', 'New morpheme after' or 'Delete morpheme', and in similar fashion as above, in the first two cases an empty column is created for being filled manually.

These manual processes can be performed also if one has chosen one of the first two bullet-point options presented above. Thus, one can at any point go back and correct mistakes, fill in more information, etc.

Note however that the only steps where you get segment splitting showing up automatically is in option 2, while option 1 just gives you the text-line in the first tier of the table.