Iramuteq DIR326.2

Importing Corpus into Iram Teec

Steps to Import the Document

• Begin by clicking the red button with a "T" at the top of the Iram Teec interface to import your corpus.

• Select the folder containing your corpus file and click "Open" to proceed with the import process.

• In the import window, set character encoding to UTF-8, which is essential for proper text processing.

• Choose Portuguese as the language of your text in the language settings and enable the default dictionary option.

• After confirming these settings, click "OK." A confirmation window should display 60 documents if successful.

Analyzing Text Segments

Understanding Document Segmentation

• The imported corpus consists of 60 documents divided into 285 segments, where each segment represents fragments of up to three lines.

• Occurrences refer to total word counts in the text, while forms indicate different classes of words present in the document.

• The next step involves generating graphs for analysis; access this feature through "Text Analysis" on the interface.

Configuring Analysis Settings

Customizing Word Classes for Analysis

• In properties settings, you can customize which word classes are included in your analysis based on relevance.

• Input zero for any word class that does not interest you (e.g., definite articles, pronouns), effectively excluding them from analysis.

• This customization simplifies data interpretation by focusing only on relevant linguistic elements before finalizing settings.

Interpreting Graphical Data

Understanding Bidimensional Graph Representation

• The generated graph is bidimensional: X-axis shows word count while Y-axis indicates frequency of each word's appearance.

Analysis of Textual Data

Understanding Specificity in Text Analysis

• The analysis begins with a focus on the frequency of specific words, emphasizing the importance of understanding their occurrence within documents.

• Users are guided to open a pre-configured window for analyzing text specificity, ensuring that their document is selected before proceeding.

• A generated table displays word occurrences across different documents, indicating how often each word appears and its relevance to each document.

• The concept of "aderência" (adherence) is introduced, explaining that higher numbers indicate greater frequency and relevance of a word in a particular document.

• This analysis allows users to discern which documents discuss certain themes more prominently, providing insights into content focus.

Contextual Analysis of Word Usage

• Users can double-click on specific words to reveal their context within the text, enhancing understanding of how terms are applied by authors or judges.

• By examining contexts where terms like "moral" are used frequently, users gain insight into judicial reasoning and decision-making processes.

• Understanding these contexts aids future legal writing by aligning arguments with established judicial language and preferences.

Classification Methodology

• Transitioning to classification analysis using the Rener method, users are instructed to click through familiar settings without needing extensive reconfiguration.

• The tool generates graphical representations based on document similarities, showcasing how texts cluster based on shared characteristics.

Insights from Graphical Representations

• The speaker reflects on personal experiences learning this analytical tool over a year, highlighting initial challenges faced in utilizing it effectively.

• The generated graphs categorize 60 judicial decisions into five groups based on textual similarity, streamlining data interpretation significantly.

Value of Automated Document Analysis

• Each group’s size is represented visually; larger bars indicate more documents sharing similar content or themes.

• Percentages within these groups illustrate the proportion of documents discussing similar topics—valuable information that would otherwise require manual sorting and reading.

Understanding Document Similarity Analysis

Hierarchical Reading of Classes

• The tool provides guidance on how to read documents, emphasizing a logical sequence rather than random access.

• It suggests starting with Class 2, then moving to Class 3, and subsequently returning to Class 1 before proceeding to Class 4.

• This structured approach highlights the interconnectedness of classes, indicating a hierarchy in document reading.

Visualizing Data with Graphs

• A button within the tool generates vertical graphs that display frequent words within each class.

• For instance, in Class 2, the most common word is "pleitear," alongside others like "alegar" and "pagamento," hinting at themes related to labor law claims.

• These frequent terms help users understand the context and content of each class more effectively.

Grouping Documents by Themes

• The analysis separates documents into groups based on shared themes but does not initially identify which documents belong to which group.

• By using specific color codes (e.g., red for one theme and green for another), users can isolate and read documents that discuss similar topics within labor law.

Conducting Similarity Analysis

• Users initiate a similarity analysis by selecting relevant words and their frequencies from a provided list.

• A graph generated from all words may appear cluttered; thus, filtering for words appearing more than a specified frequency (e.g., greater than 20 times) is recommended.

Finalizing Insights from Judicial Decisions

• The analysis focuses on extracting insights from judicial decisions regarding labor law issues such as overtime claims.

Analysis of Judicial Sentences and Graphical Representation

Downloading Judicial Sentences

• The speaker discusses the process of downloading judicial sentences from a tribunal's website to create a corpus for analysis.

• Emphasizes the ability to analyze judges' profiles or specific branches of law, highlighting the versatility in data extraction.

Understanding Graphical Connections

• Introduces a graph showing connections between words, where thicker lines indicate stronger relationships among terms.

• Notes that the word "reclamante" (claimant) appears frequently alongside "pagamento" (payment), suggesting a strong correlation in legal texts.

Payment Contextualization

• Discusses various types of payments related to claims, such as additional payments, indemnities, and moral damages.

• Refers to 60 judicial decisions previously analyzed, indicating an ongoing exploration of these cases.

Enhancing Graphical Analysis

• The speaker encourages improving the graphical representation by preserving previous configurations for ease of analysis.

• Explains how adjusting settings can help visualize connections more clearly without needing to reconfigure everything.

Grouping Data Insights

• Introduces a feature that groups related terms within the graph based on their connections and contexts.

• Highlights how different groups are formed around concepts like payment types and company structures (e.g., LTDA).

Final Thoughts on Data Visualization

• Suggestion to disable certain functions while maintaining community grouping for clearer insights into data relationships.

Graph Analysis Techniques

Introduction to Graph Configuration

• The speaker emphasizes the importance of preserving previous settings when generating a new graph by clicking a specific button.

• After selecting "Communities," the graph displays groups with distinct colors, enhancing visual differentiation.

Understanding Graph Features

• The final configuration involves disabling certain options to focus solely on community representation, which simplifies the analysis.

• The thickness of connections in the graph indicates relationship intensity; thicker lines represent stronger relationships while thinner lines indicate weaker ones.

Word Cloud Analysis

• Transitioning to text analysis, the speaker introduces word clouds and explains how to set dimensions for better visualization (height: 400, width: 400).

• A maximum of 100 words is recommended for clarity; too many words can lead to a cluttered display.

Interpreting Word Clouds

• The size of words in the cloud reflects their frequency in the analyzed documents; larger words appear more frequently.

• Specific terms related to labor law are highlighted as examples, demonstrating how word clouds can reveal thematic relevance in texts.

Final Steps and Submission Guidelines

• Participants are instructed to select graphs for their analyses and submit them by Monday of the following week.

• Emphasis is placed on analyzing five key graphs for sufficient understanding before moving forward with written analyses.

Discussion on Team Rivalries

Light-hearted Banter about Football Teams

• A humorous exchange occurs regarding football teams, particularly Corinthians and Flamengo, showcasing regional rivalries.

Cultural Commentary on Fan Identity

• Discussion touches upon fan identities and perceptions among supporters of different teams, highlighting passionate affiliations.

Recipe for Improvement

Introduction to the Session

• The speaker humorously addresses a colleague, mentioning their shared interest in football (Flamengo), indicating a light-hearted atmosphere.

• The speaker expresses concern about managing their responsibilities as a teacher while being on medical leave, highlighting the importance of student progress.

Preparing for Analysis

• The speaker discusses the need to select specific graphs for analysis, emphasizing organization and clarity in data presentation.

• Instructions are given on how to visualize graphs better by adjusting icon sizes within folders, enhancing user experience during analysis.

Graph Utilization

• The first graph is introduced; it separates documents and is intended for use in Word. The process of copying and pasting is outlined clearly.

• Emphasis is placed on finding additional necessary graphs within the same folder, reiterating the copy-paste method for transferring data into Word.

Navigating Folders

• Transitioning to another folder named "sim TXT," which contains similarity analysis results. Instructions are provided on how to view these files effectively.

• A specific graph from this folder is deemed sufficient for transfer to Word, reinforcing efficiency in selecting relevant data.

Final Steps in Data Transfer