Tools Home : HTML Tools : Find Text — Collocation

Click here to show HTML tools HTML Tools

Click here to expand XML tools XML tools

Click here to expand plain text tools Plain Text Tools

Click here to expand other tools Other tools

 Beta tools
 Add Tools Demo
 Manual
 About

Find Text — Collocation
?
Summary

This tool searches for a word or pattern and places it into a user-defined context of words, sentences, lines or paragraphs within an HTML document located either at a user-specified web address or uploaded from the user’s files. The results can be sorted alphabetically, by frequency or by Z-score (a measure of how far and in what direction that word deviates from it’s distribution’s mean in units of its standard deviation)

Note 1: When the set context is large, it is likely that the word or pattern searched will appear more than once in a given context. In this instance, the word or pattern will be counted more than once, decreasing the accuracy of the count and Z-score. For this reason, the best results are returned from small contexts.

Note 2: When 'Paragraphs' is selected as context, any element name entered by the user will be overwritten by the 'p' (<p></p>) tag. As not all HTML texts use the paragraph tag to identify paragraphs, this may cause some problems with the tool’s results. If so, use a context of word, line or sentence instead.

Please click the ? buttons at the bottom right of each set of options for more information on that set.

For further information on this tool, please see the TADA Wiki's Collocation entry here. A glossary of terms is also available here.

Walkthrough

To generate collocates for text found within http://www.w3.org/, extract text from between <body> and </body>, limit the results to strings of text that contain 'web', generate a list of the ten words appearing before and after each instance, sort the results alphabetically and display it as HTML:
  1. Source text
    1. Enter ‘http://www.w3.org/’ into the ‘URL’ field.
  2. Subtext limited to
    1. Enter ‘body’ in the ‘Elements’ field.
  3. What to find
    1. Enter ‘web’ into the ‘Word/Pattern’ field.
  4. Context for concordance
    1. Set the ‘Context’ drop list to ‘Words’.
    2. Enter ‘10’ into the ‘Context length’ field.
  5. Results
    1. Set the ‘Sort’ drop menu to ‘Co-occurring words alphabetically’.
    2. Set the ‘Display as’ drop menu to ‘HTML’.
  6. Click the ‘Submit’ button to process the text.
*
» Source text
  Example: http://taporware.ualberta.ca/einstein-bio.html

?
Summary

This section determines the source of the document you wish the tool to process. HTML can be obtained either from a web address or by uploading a file.

Fields

Source URL
To use content from a web page, enter a full web address (URL) in the field provided. Copy and paste from your browser’s address bar for best results.

Local file
To upload an HTML (.html) file from your computer, choose ‘Local file,’ click ‘Browse,’ and select the file you wish to use from your directory.
» Subtext limited to
(separate multiple elements with a `,')
?
Summary

This section determines which HTML tags to extract text from.

Fields

Elements
Use this field to specify which HTML tag(s) to extract text from. Multiple tags must be separated by commas (ex: 'p, h1, h2'). This field defaults to 'body'.
*
» What to find
?
Summary

This section allows the source text to be searched for words, phrases or patterns.

Fields

Word/Pattern
Use this field to filter results based on the word, phrase or pattern (regular expression) entered here. To search multiple words without the tool treating them as a phrase, separate words with commas (Ex: red, orange, purple). To search for a phrase, enter it as it appears in the text - quotation marks are not needed (Ex: wine-dark sea, not "wine-dark sea"). Unix-style searching may also be used.

Get Synonyms
If searching a word, this button generates a list of synonyms in a new window. To search the original word plus synonymns, copy the comma-separated list in the text box of the new window and paste it into the Word/Pattern field. Note: Concordance only.

Exclude modified Glasgow Stop Words
Choose this option to filter the modified Glasgow Stop Words list out of the final list Note: Collocation only.
» Context for concordance
?
Summary

This section allows the user to define the context type (words, lines, sentences or paragraphs), and how many of that type (context length) to show on either side of each instance of the word or pattern.

Fields

Context

Context can be set to one of four options from the drop menu: words, lines, sentences and paragraphs.

Words
This option places the search term or pattern in context by a specified number of words.

Lines
This option places the search term or pattern in context by a specified number of lines.

Sentences
This option places the search term or pattern in context by a specified number of sentences.

Paragraphs
This option places the search term or pattern in context by a specified number of paragraphs.

Context Length
This drop menu allows users to enter how many words, lines, sentences or paragraphs to show before and after the word or pattern. For example, if the context length is 5 words, every instance of the search term or pattern will be shown surrounded by a total of ten words - the five that precede it, and the five that follow it.
» Results
?
Summary

This section allows users to choose how the results will be formatted, and whether to display it in a new browser window.

Fields

Sort
This drop-down lists enables users to choose from several sort options: Co-occurring words by frequency, co-occurring words alphabetically, and co-occurring words by z-score.

Display as
This drop-down lists enables users to choose from several output formats: HTML, XML text in HTML, XML tree, and Tab delimited text.

Open results in new window
Check this box to display the results in a new window or browser tab. This option is selected by default. Some pop-up blockers may prevent a new window from being opened; if so, un-check the box to open the results in the same window instead.
`*' indicates a required field

 

 

TAPoRware Project, McMaster University,