Update metric description in README.md

README.md

@@ -1,23 +1,53 @@
 ---
-title: Metric Template 1
+title: ISCO-08 Hierarchical Accuracy Measure
 datasets:
 -  ICILS/multilingual_parental_occupations
 tags:
 - evaluate
 - metric
-description: "TODO: add a description here"
+description: "The ISCO-08 Hierarchical Accuracy Measure is an implementation of the measure described in [Functional Annotation of Genes Using Hierarchical Text Categorization](https://www.researchgate.net/publication/44046343_Functional_Annotation_of_Genes_Using_Hierarchical_Text_Categorization) (Kiritchenko, Svetlana and Famili, Fazel. 2005) applied with the ISCO-08 classification scheme by the International Labour Organization."
 sdk: gradio
 sdk_version: 3.19.1
 app_file: app.py
 pinned: false
 ---
 
-# Metric Card for Metric Template 1
+# Metric Card for ISCO-08 Hierarchical Accuracy Measure
 
 ***Module Card Instructions:*** *Fill out the following subsections. Feel free to take a look at existing metric cards if you'd like examples.*
 
 ## Metric Description
-*Give a brief overview of this metric, including what task(s) it is usually used for, if any.*
+The ISCO-08 Hierarchical Accuracy Measure is an implementation of the measure described in [Functional Annotation of Genes Using Hierarchical Text Categorization](https://www.researchgate.net/publication/44046343_Functional_Annotation_of_Genes_Using_Hierarchical_Text_Categorization) (Kiritchenko, Svetlana and Famili, Fazel. 2005) and adapted for the ISCO-08 classification scheme by the International Labour Organization.
+
+The measure accounts for the hierarchical structure of the ISCO-08 classification scheme by rewarding partial correctness in classification, penalizing distant errors, and applying higher penalties for higher-level errors. It is designed to evaluate the accuracy of predictions within a hierarchical occupation classification system, such as the ISCO-08 scheme, which organizes occupations into a tree structure with four levels: major group, sub-major group, minor group, and unit group.
+
+1. **Partial correctness in classification**: Misclassification within the same major group (e.g., within "Professionals") is penalized less than misclassification across major groups (e.g., from "Professionals" to "Managers"). This acknowledges the similarity in skill level and specialization within major groups.
+
+2. **Penalization of distant errors**:
+ 2.1 Correct classification into a more specific group (e.g., "Medical doctors" within "Health professionals") is preferred over a broader classification (e.g., the entire "Health professionals" group).
+ 2.2 Conversely, incorrect classification into a more specific but unrelated group (e.g., "Nursing and midwifery professionals" when the correct classification is "Engineering professionals") is penalized more heavily than an incorrect but broader classification.
+
+3. **Higher penalties for higher-level errors**: Misclassifications at the major group level are more severely penalized than errors within sub-major, minor, or unit groups, reflecting the greater disparity in skill level and specialization.
+
+4. **Extension with ancestors in the hierarchy**: For the hierarchical precision ($hP$) and recall ($hR$) calculations, an occupation's classification is extended to include not only its specific group but also the parent groups to which it belongs, excluding the root level. This allows for an evaluation that recognizes partial correctness based on the occupation's position within the ISCO-08 hierarchy.
+
+### Original description
+
+1. The measure gives credit to partially correct classification, e.g. misclassification into node $I$ (ISCO unit group "1120") when the correct category is $G$ (ISCO unit group "1111") should be penalized less than misclassification into node $D$ (e.g., ISCO unit group "1211") since $I$ is in the same subgraph (ISCO sub-major group "11") as $G$ and $D$ is not.
+2. The measure punishes distant errors more heavily:
+    1. the measure gives higher evaluation for correctly classifying one level down compared to staying at the parent node, e.g. classification into node $E$ (ISCO minor group "111") is better than classification into its parent $C$ (ISCO sub-major group "11") since $E$ is closer to the correct category $G$; and
+    2. the measure gives lower evaluation for incorrectly classifying one level down comparing to staying at the parent node, e.g. classification into node $F$ (ISCO unit group "1120") is worse than classification into its parent $C$ since $F$ is farther away from $G$.
+\n
+The features described are accomplished by pairing hierarchical variants of precision ($hP$) and recall ($hR$) to form a hierarchical F1 (hF_β) score where each sample belongs not only to its class (e.g., a unit group level code), but also to all ancestors of the class in a hierarchical graph (i.e., the minor, sub-major, and major group level codes).\n
+\n
+Hierarchical precision can be computed with:\n
+$hP = \frac{| \v{C}_i ∩ \v{C}^′_i|}  {|\v{C}^′_i |} = \frac{1}{2}$\n
+\n
+Hierarchical recall can be computed with:\n
+$hR = \frac{| \v{C}_i ∩ \v{C}^′_i|}  {|\v{C}_i |} = \frac{1}{2}$\n
+\n
+Combining the two values $hP$ and $hR$ into one hF-measure:\n
+hF_β = \frac{(β^2 + 1) · hP · hR}{(β^2 · hP + hR)}, β ∈ [0, +∞)
 
 ## How to Use
 *Give general statement of how to use the metric*