Upload README.md

README.md

@@ -35,14 +35,14 @@ widget:
   example_tite: "Unlabeled 6"
 ---
 
-# WRAP -- A Content Management System for Twitter
+# WRAP -- A TACO-based Classifier For Inference and Information-Driven Argument Mining on Twitter
 
 Introducing WRAP, an advanced classification model built upon `AutoModelForSequenceClassification`, designed to identify tweets belonging to four 
 distinct classes: Reason, Statement, Notification, and None of the [TACO dataset](https://anonymous.4open.science/r/TACO).
 Designed specifically for extracting information and inferences from Twitter data, this specialized classification model utilizes 
 [WRAPresentations](https://huggingface.co/TomatenMarc/WRAPresentations), from which WRAP acquires its name.
-WRAPresentations is an advancement of the [BERTweet-base](https://huggingface.co/vinai/bertweet-base) architecture, whose pre-training was 
-extended on augmented tweets using contrastive learning.
+WRAPresentations is an advancement of the [BERTweet-base](https://huggingface.co/vinai/bertweet-base) architecture, whose embeddings were 
+extended on augmented tweets using contrastive learning for better encoding inference and information in tweets.
 
 ## Class Semantics
 
@@ -63,7 +63,7 @@ or quotation, and thus reveals the author’s motivation *to try to understand a
 In its entirety, WRAP can classify the following hierarchy for tweets:
 
 <div align="center">
-  <img src="https://github.com/TomatenMarc/public-images/raw/main/Component_Space_WRAP.svg" alt="Component Space" width="100%">
+  <img src="https://github.com/TomatenMarc/public-images/raw/main/Argument_Tree.svg" alt="Component Space" width="100%">
 </div>
 
 ## Usage
@@ -85,7 +85,7 @@ prediction = pipe("Huggingface is awesome")
 print(prediction)
 ```
 
-<a href="https://github.com/TomatenMarc/TACO/blob/main/notebooks/classifier_cv.ipynb">
+<a href="https://anonymous.4open.science/r/TACO/notebooks/classifier_cv.ipynb">
     <blockquote style="border-left: 5px solid grey; background-color: #f0f5ff; padding: 10px;">
         Notice: The tweets need to undergo preprocessing before classification.
     </blockquote>
@@ -113,7 +113,7 @@ Additionally, the category and class distribution of the dataset TACO is as foll
 
 <p>
     <blockquote style="border-left: 5px solid grey; background-color: #f0f5ff; padding: 10px;">
-        Notice: Our training involved WRAP to forecast class predictions, where the categories (Information/Inference) represent class aggregations 
+        Notice: Our training involved WRAP to forecast class predictions, where the categories (information/inference) represent class aggregations 
 based on the inference or information component.
     </blockquote>
 <p>
@@ -161,7 +161,7 @@ In total, the WRAP classifier performs as follows:
 | Macro-F1    | Inference | Information | Multiclass |
 |-------------|-----------|-------------|------------|
 | In-Topic    | 87.71%    | 85.34%      | 75.80%     |
-| Cross-Topic | 86.71%    | 84.59%      | 73.92%     |
+| Cross-Topic | 86.71%    | 84.58%      | 73.92%     |
 
 ### Classification