Merge pull request #632 from PrimozGodec/scorer-widget

[ENH] Score documents widget

doc/widgets/score-documents.md

@@ -0,0 +1,25 @@
+Score Documents
+===============
+
+Scores documents based on word appearance.
+
+**Inputs**
+
+- Corpus: A collection of documents
+- Words: A table with at least one column with words
+
+**Outputs**
+
+- Corpus: A collection of documents with scores
+
+**Score Document** assigns scores to each document based on words at the input. The score with the selected scoring method is calculated for each word and aggregated together with the selected aggregation function on the document level.
+
+![](images/Score-Documents.png)
+
+1. Select one or more scoring methods for words:
+   - **Word frequency**: The count of the frequency of a word in the text.  
+   - **Word ratio**: Indicate the appearance of the word in the document.
+   - **Similarity**: The cosine similarity between document embedding and word embedding.
+2. Select aggregation function to aggregate word scores in document scores.
+3. Filter documents based on the document title in the first column
+4. The table with the document titles in the first column and scores for in other columns.

orangecontrib/text/vectorization/document_embedder.py

@@ -6,7 +6,7 @@
 import json
 import sys
 import warnings
-from typing import Tuple, Any, Optional
+from typing import Tuple, Any, Optional, Union, List
 import numpy as np
 
 from Orange.misc.server_embedder import ServerEmbedderCommunicator
@@ -93,13 +93,13 @@ def __init__(self, language: str = 'en',
                                          embedder_type='text')
 
     def __call__(
-        self, corpus: Corpus, processed_callback=None
-    ) -> Tuple[Corpus, Corpus]:
+        self, corpus: Union[Corpus, List[List[str]]], processed_callback=None
+    ) -> Union[Tuple[Corpus, Corpus], List[Optional[List[float]]]]:
         """Adds matrix of document embeddings to a corpus.
 
         Parameters
         ----------
-        corpus : Corpus
+        corpus : Corpus or list of lists
             Corpus on which transform is performed.
 
         Returns
@@ -114,12 +114,15 @@ def __call__(
         ValueError
             If corpus is not instance of Corpus.
         """
-        if not isinstance(corpus, Corpus):
-            raise ValueError("Input should be instance of Corpus.")
+        if not isinstance(corpus, (Corpus, list)):
+            raise ValueError("Input should be instance of Corpus or list.")
         embs = self._embedder.embedd_data(
-            list(corpus.ngrams),
+            list(corpus.ngrams) if isinstance(corpus, Corpus) else corpus,
             processed_callback=processed_callback)
 
+        if isinstance(corpus, list):
+            return embs
+
         dim = None
         for emb in embs:  # find embedding dimension
             if emb is not None: