v 2.0 User`s Manual
Contents
1. lt cnode gt lt cnode name CERTAIN1 gt lt pnode name CERTAIN gt lt pnode gt lt cnode gt lt cnode name CERTAIN2 gt lt pnode name CERTAIN gt lt pnode gt lt cnode gt lt dictionary gt Again this dictionary is on the simple side But it has five categories with a varying number of words under each This dictionary is built just to test or demonstrate Lexicoder To run the dictionary let s first select the Article Cleaner Select Article Cleaner from the Select Processors drop down menu and click on the Add button The Article Cleaner should now appear in the processors winder Let s also get a count of the total number of words in the article Select Word Counter from the Select Processors drop down menu and click on the Add button Now select Dictionary Counter from the Select Processors drop down menu and click on the Add button The dictionary counter should now show up the processors window The Lexicoder window now looks as follows Lexicoder User s Manual August 2011 version Page 8 KoKo Lexicoder File Help Select Source Canzo STAT Tab Delimited File z Load Source ObamaExample txt ObamaExample txt loaded sucess a Processor Article Cleaner v2 0 ad Select Processors Processor Word Counter v1 0 add Dictionary Counter v2 0 w Article Cleaner v2 0 Processor Dictionary Counter v2 0 Word Counter v1 0 Add Dictionary Counter v2 0 Configure Remove Clear All Tab2. lt pnode name black gt lt pnode name brown gt lt pnode name red gt lt cnode gt lt dictionary gt Now if your text looked like this The quick brown fox jumps over the lazy dog Then the results of the analysis would look like this Article ID Body Animals Colours 1 The quick brown fox jumps over the lazy 2 1 dog There are two animal words fox dog and one color brown 2E First Mentions v1 0 This module uses the same kind of dictionary files as the Dictionary Counter but in this case it captures the character at which a given word begins The processor is useful in instances in which the research wants to know for instance whether the article mentions Democrats first or Republicans first The processor returns the number of characters in the text up to and including the first character for Lexicoder User s Manual August 2011 version Page 5 the first instance in which in a word included in a given dictionary appears Smaller numbers then indicate that the word appears earlier in the article 2F Sentence Proximity Analyzer v1 0 This module uses the same kind of dictionary files as the Dictionary Counter but in this case it counts the occurrences of words specified in one dictionary that appear in the same sentence as words specified in another dictionary The analyzer requires some careful pre processing of the data because sentences are identified usi
3. a xX lexi v20 User s Manual Lexicoder is a Java based multi platform tool for automated content analysis To install Lexicoder on your computer PC or Mac you need only unzip the Lexicoder zip file It can be saved in any location on your computer To start Lexicoder you need to double click on the Lexicoder jar file Lexicoder will typically take a few moments to start If it does not start or reports an error please be sure that you have the latest version of Java installed on your computer Once it s up and running the Lexicoder interface is very simple eoo Lexicoder File Options Help Console Output Load Source 5 Select Processors Article Cleaner v1 0 v Add Configure Remove Clear All Load Sink Process x 4 Il gt There are four steps to analyze data in Lexicoder each represented by one of the Button on the left of the interface Load Source Select Processors Add Load Sink and Process The Console Output window confirms each of these processing steps We ll describe how to use Lexicoder here first by describing briefly each of the four steps in the process and then by walking step by step through a content analysis of an example dataset using an example dictionary Lexicoder User s Manual August 2011 version Page 2 Section 1 The Four Steps 1 Load Source Select the file that you wish to an
4. Dem Lexicoder User s Manual August 2011 version Page 3 M P MP M P P MPP Dr Dr s s replaced with a space replaced with a space replaced with a space replaced with a space replaced with a space replaced with a space replaced with a space replaced with a space replaced with a space replaced with a space ay A N Some text will require much more pre processing This can be accomplished with any text browser that includes a find and replace mechanism Note that this is the only module in this release that actually changes the contents of the article which is processed Thus the ordering of this particular module matters Putting for example the dictionary counter before or after the word counter will not impact the results for those modules placing the article cleaner before or after the dictionary counter however will result in different outcomes 2B Word Counter v1 0 The word counter provides a simple count of the number of words in the text 2C Article Stemmer v1 0 This is the Porter Stemming Algorithm developed by Martin Porter Information is available at http tartarus org martin PorterStemmer 2D Dictionary Counter v2 0 This is the most critical module for most users it counts the occurrences of words specified in a dictionary The dictionary must meet certain specifications see samples below It must be
5. formatted in XML and it must be two leveled If the dictionary is not properly formatted Lexicoder will not be able to work with it and may not be able to tell you When designing a dictionary it is important to keep the following processing details in mind The Dictionary Counter goes through each category in the dictionary sequentially first level entries Lexicoder User s Manual August 2011 version Page 4 It goes through each word or phrase in the category and checks to see if it exists in the article It counts and adds up each existence of the word or phrase in the text As it counts each existence of the word or phrase is removed from future consideration by the module though all content will remain for other modules This is to improve efficiency as well as to properly handle phrases Thus if a word or phrase occurs in two categories the dictionary counter will only count it as belonging to the first category By way of example image that you have dictionary that captures two categories animals and colors The dictionary in the correct xml format for Lexicoder might look like this lt xml version 1 0 encoding UTF 8 standalone no gt lt dictionary style Lexicoder name Test Dictionary gt lt cnode name Animals gt lt pnode name fox gt lt pnode name cow gt lt pnode name dog gt lt cnode gt lt cnode name Colours gt
6. Delimited File mal Load Sink Process ly 4 il gt The source is listed after the Load Source button the processors are listed in the processor window and each step has been identified in the Console Output window The Article Cleaner does not need to be configured but the Dictionary Counter does So select Dictionary Counter v2 0 in the processor window so it is highlighted and click on the Configure button at the bottom left of the processor window A window will pop up and you can select the dictionary file you want to use For this example select ObamaDict Icd Once you have selected the dictionary a pop up window will ask if you want the dictionary to be case senstive or not If you select the case sensitive option a dictionary entry of OBAMA will not match text that reads Obama For some searches this is desirable In most cases however not case sensitive is most appropriate In this case select No Now you need to select the file into which results will be saved With Tab Delimited File selected click on the Load Sink button Name your file and save it anywhere you like Then click on Process You will get a pop up window when your analysis is completed The results can then be opened using any text editing software or database software such as Excel The file will include five columns the original ID number in the first column and then the five dictionary categories in the subsequent
7. ObamaDict Icd dictionary Both are available at lexicoder com ObamaExample txt includies President Obama s 2009 inauguration speech Each entry is a separate paragraph from the speech there are 36 entires in all Note that this text has just two columns ID and Body and that they are separated by a tab So long as a tab delimited file is saved in this format it can be loaded into Lexicoder To start then let s open the ObamaExample txt dataset With Tab Delimited File in the drop down menu under Select Source click on the Load Source button and select the file The dictionary which can be opened and editing using any text processor is as follows lt xml version 1 0 encoding UTF 8 standalone no gt Lexicoder User s Manual August 2011 version Page 7 lt dictionary style Lexicoder name ObamaDictionary gt lt cnode name POLITE gt lt pnode name HUMBLE gt lt pnode gt lt pnode name GRATEFUL gt lt pnode gt lt pnode name MINDFUL gt lt pnode gt lt cnode gt lt cnode name BAD gt lt pnode name CRISIS gt lt pnode gt lt pnode name FEAR gt lt pnode gt lt pnode name CONFLICT gt lt pnode gt lt pnode name DISCHORD gt lt pnode gt lt pnode name DOUBT gt lt pnode gt lt cnode gt lt cnode name GOOD gt lt pnode name HOPE gt lt pnode gt lt pnode name UNITY gt lt pnode gt lt pnode name FREEDOM gt lt pnode gt lt pnode name LIBERTY gt lt pnode gt
8. alyze This file should be plain tab delimited text It should have one column with case IDs labelled ID That ID will be saved alongside results and you will need that ID in order to merge those results with your original datafile It should also have one column labelled Body this is the column containing the data that Lexicoder will analyze 2 Select Processors This is the stage at which you select the various processors you would like to use on your data There are several options listed in the drop down menu and they can be added in any order you wish The current version of Lexicoder includes six processors available through the drop down window as follows Select Processors Article Cleaner v1 0 v Article Cleaner v1 0 Word Counter v1 0 Article Stemmer v1 0 Dictionary Counter v2 0 First Mentions v1 0 Sentence Proximity Analyze 2A Article Cleaner v1 0 This is typically the first processor in any analysis The cleaner goes through your dataset and by replacing and re shuffling punctuation marks tries to reduce errors in subsequent analyses Commas decimals and periods question marks dashes and the like can affect word counts and the identification of sentences for instance The following is a list of the conversions that the current article cleaner v2 0 performs Original Text Clean Text Mr Mr Mrs Mrs U S US U S A USA U S S R USSR Rep Rep Dem
9. columns Lexicoder User s Manual August 2011 version Page 9 Results for this particular file are as follows though the columns may not be saved in exactly this order ID Word Count POLITE BAD GOOD CERTAINICERTAIN2 1 3 0 0 0 0 0 2 51 3 0 0 0 0 3 82 0 0 0 0 0 4 115 0 1 0 0 0 5 45 0 2 0 0 0 6 42 0 0 0 0 0 7 21 0 2 2 0 0 8 34 0 0 0 0 0 9 80 0 0 0 0 0 10 103 0 0 1 0 0 11 20 0 0 0 0 0 12 23 0 0 0 0 0 13 19 0 0 0 0 0 14 49 0 0 0 0 0 15 100 0 1 0 0 0 16 134 0 0 0 0 0 17 58 0 0 0 0 0 18 141 0 0 0 0 0 19 113 0 1 2 0 0 20 132 0 0 0 0 0 21 79 0 0 0 0 0 22 136 0 0 0 0 0 23 122 0 0 0 0 0 24 99 0 1 0 0 0 25 80 0 0 0 0 0 26 109 1 0 1 0 0 27 94 0 0 0 0 0 28 148 0 0 0 0 0 29 9 0 0 0 0 0 30 20 0 0 0 1 0 31 63 0 0 1 0 0 32 89 0 1 0 0 0 33 42 0 0 1 0 0 34 103 0 0 2 0 0 35 5 0 0 0 0 0 36 8 0 0 0 0 0 There are 3 words from the POLITE category in the second paragraph of the speech That paragraph is as follows I stand here today humbled by the task before us grateful for the trust you have bestowed mindful of the sacrifices borne by our ancestors thank President Bush for his service to our nation as well as the generosity and cooperation he has shown throughout this transition Lexicoder User s Manual August 2011 version Page 10 Note that the dictonary counter captures the word humble even though it has a suffix in this line This is an important feature of the dictionary counter If you w
10. icoder User s Manual August 2011 version Page 6 1 The quick brown fox jumps over the lazy dog 1 At least I though it was brown it might have been red Note that the number of colors counted is just one as it was earlier The additional sentences include other color words brown again and red but there is no animal in those sentences Colors words are only identified when they co occur with animal words 3 Load Sink When you click on Load Sink a window will pop up and you will name the files into which you would like results to be stored There are no particular rules here data will be stored as plain tab delimited text no matter what name you use here It may help to use the txt or tab suffix here depending on the software you will be using to open up the resulting dataset Microsoft Excel Apple Numbers Filemaker and StatTransfer will be able to open the output without any difficulty 4 Process Click on the Process button Sit back and wait You will see a progress panel and Lexicoder will let you know when the data are ready The processing is of course the easiest step in a content analysis getting and formatting the content and designing your dictionary will take far more time See the Lexicoder website for citations and links to dictionaries and work on building dictionaries Section 2 An Example Let s walk through an analysis of the ObamaExample txt file using the
11. ish to avoid suffixes you should leave a space after humble in the quotation marks that is you should use humble in the dictionary The categories CERTAIN1 and CERTAIN1 are designed to show this as well though using the beginning of the word Note that CERTAIN1 includes the word certain which CERTAIN2 includes the word certain with a space in front of it Line 30 is This is the source of our confidence the knowledge that God calls on us to shape an uncertain destiny CERTAIN1 counts the word certain even though it is preceded by un CERTAIN2 does not count the word certain unless it is preceded by a space The process of identifying the many prefixes and suffixes or words and including the appropriate forms while excluding the others can of course be very complicated It is important that you build your dictionary keeping in mind then the use or not of spaces as well as the fact that the dictionary counts words in the order in which they appear in the dictionary and then sets them aside Section 3 Final Comments If you want to practice more there are additional files available at lexicoder com In particular there is USinaugurationsa1949 2009 txt This file includes predictably all inaugural addresses since 1949 Here the cases are the addresses themselves they are not divided by paragraph The Lexicoder Topic Dictionary LTDv2 lcd is an early version of a topic dictionary designed to use w
12. ith Lexicoder It is a much more complicated dictionary than the ones we have used thus far It is by no means a finished product but it is a useful file for practicing and also a good template as you design your own dictionary Other dictionaries are available at lexicoder com as well including the Lexicoder Sentiment Dictionary a dictionary designed to capture the sentiment in political texts Lexicoder was programmed by Mark Daku and developed by Lori Young and Stuart Soroka at McGill University Comments and queries are very welcome at lexicoderOme com
13. ng periods so all the extra periods in Mr for instance have to be removed before you can run this module reliably See the Article Cleaner above Image you are using a similar dictionary to the one used above though this time just for colors lt xml version 1 0 encoding UTF 8 standalone no gt lt dictionary style Lexicoder name Color Dictionary gt lt cnode name Colours gt lt pnode name black gt lt pnode name brown gt lt pnode name red gt lt cnode gt lt dictionary gt The Sentence Proximity Analyzer asks you to define a second dictionary the dictionary the words with which animals and colors need to co occur in order to be counted Your second dictionary is as follows it identifies animals lt xml version 1 0 encoding UTF 8 standalone no gt lt dictionary style Lexicoder name Animal Dictionary gt lt cnode name Animals gt lt pnode name fox gt lt pnode name cow gt lt pnode name dog gt lt cnode gt lt dictionary gt Now if your text looked like this The quick brown fox jumps over the lazy dog At least I though it was brown it might have been red The result of searching for the first set of words colors conditional on their co occurrence with the second set of words animals is Article ID Body Colors Lex
Download Pdf Manuals
Related Search