diff --git a/l3/TM-Lab3.ipynb b/l3/TM-Lab3.ipynb
index ce9a0ae66b8e5a78767290517ffe5ab1c5d5dcba..6bf502bbaf6c675c49462691e825c3c2612df510 100644
--- a/l3/TM-Lab3.ipynb
+++ b/l3/TM-Lab3.ipynb
@@ -37,7 +37,7 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 303,
"metadata": {
"deletable": false,
"editable": false,
@@ -82,7 +82,7 @@
},
{
"cell_type": "code",
- "execution_count": 4,
+ "execution_count": 304,
"metadata": {
"deletable": false,
"editable": false,
@@ -128,7 +128,7 @@
},
{
"cell_type": "code",
- "execution_count": 5,
+ "execution_count": 305,
"metadata": {},
"outputs": [
{
@@ -220,7 +220,7 @@
"4 Brussels "
]
},
- "execution_count": 5,
+ "execution_count": 305,
"metadata": {},
"output_type": "execute_result"
}
@@ -262,7 +262,7 @@
},
{
"cell_type": "code",
- "execution_count": 6,
+ "execution_count": 306,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -312,7 +312,7 @@
},
{
"cell_type": "code",
- "execution_count": 7,
+ "execution_count": 307,
"metadata": {
"deletable": false,
"editable": false,
@@ -350,7 +350,7 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": 308,
"metadata": {
"deletable": false,
"editable": false,
@@ -419,7 +419,7 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": 309,
"metadata": {
"deletable": false,
"editable": false,
@@ -464,7 +464,7 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": 310,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -498,7 +498,7 @@
},
{
"cell_type": "code",
- "execution_count": 25,
+ "execution_count": 311,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -552,7 +552,7 @@
},
{
"cell_type": "code",
- "execution_count": 27,
+ "execution_count": 312,
"metadata": {},
"outputs": [
{
@@ -575,7 +575,7 @@
},
{
"cell_type": "code",
- "execution_count": 47,
+ "execution_count": 313,
"metadata": {
"deletable": false,
"editable": false,
@@ -630,7 +630,7 @@
},
{
"cell_type": "code",
- "execution_count": 48,
+ "execution_count": 314,
"metadata": {
"deletable": false,
"editable": false,
@@ -692,7 +692,7 @@
},
{
"cell_type": "code",
- "execution_count": 49,
+ "execution_count": 315,
"metadata": {
"tags": [
"solution"
@@ -16062,7 +16062,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 323,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -16092,8 +16092,22 @@
" triples consisting of the sentence id, start position, and end\n",
" position of each span.\n",
" \"\"\"\n",
- " # YOUR CODE HERE\n",
- " raise NotImplementedError()"
+ " blacklist = ['DATE', 'CARDINAL', 'TIME', 'QUANTITY', 'PERCENT', 'ORDINAL']\n",
+ "\n",
+ " ner = nlp.get_pipe(\"ner\")\n",
+ "\n",
+ " for row in df.itertuples():\n",
+ " sentence = row[2]\n",
+ " doc = nlp(sentence)\n",
+ "\n",
+ " for ent in doc.ents:\n",
+ " if ent.label_ in blacklist:\n",
+ " continue\n",
+ "\n",
+ " if ent.text.startswith(\"the \") or ent.text.startswith(\"The \"):\n",
+ " ent.start += 1\n",
+ " \n",
+ " yield row[1], ent.start, ent.end"
]
},
{
@@ -16109,21 +16123,31 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 324,
"metadata": {
"tags": [
"solution"
]
},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Precision: 0.875, Recall: 0.799, F1: 0.835\n"
+ ]
+ }
+ ],
"source": [
"scores_improved = evaluation_scores(spans_dev_gold, set(pred_spans_improved(df_dev)))\n",
- "print_evaluation_scores(scores_improved)"
+ "print_evaluation_scores(scores_improved)\n",
+ "\n",
+ "# Previous score: Precision: 0.551, Recall: 0.775, F1: 0.644"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 325,
"metadata": {
"deletable": false,
"editable": false,
@@ -16139,7 +16163,20 @@
"task": false
}
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div class=\"alert alert-success\"><strong>Checks have passed!</strong></div>"
+ ],
+ "text/plain": [
+ "<IPython.core.display.HTML object>"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"assert scores_improved[-1] > .8, \"F1-score should be above 0.8\"\n",
"success()"
@@ -16156,7 +16193,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 326,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -16182,8 +16219,19 @@
" A *new* data frame with the same layout as `df`, but containing\n",
" the predicted start and end positions for each token span.\n",
" \"\"\"\n",
- " # YOUR CODE HERE\n",
- " raise NotImplementedError()"
+ "\n",
+ " sentence_dict = dict(zip(df['sentence_id'], df['sentence']))\n",
+ "\n",
+ " rows = []\n",
+ "\n",
+ " for index, (sentence_id, beg, end) in enumerate(pred_spans_improved(df)):\n",
+ " sentence = sentence_dict.get(sentence_id)\n",
+ " rows.append({'sentence_id': sentence_id, 'sentence': sentence, 'beg': beg, 'end': end})\n",
+ "\n",
+ " new_data_frame = pd.DataFrame(rows, columns=['sentence_id', 'sentence', 'beg', 'end'])\n",
+ "\n",
+ " return new_data_frame\n",
+ "\n"
]
},
{
@@ -16197,7 +16245,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 327,
"metadata": {
"deletable": false,
"editable": false,
@@ -16216,7 +16264,179 @@
"solution"
]
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>sentence_id</th>\n",
+ " <th>sentence</th>\n",
+ " <th>beg</th>\n",
+ " <th>end</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>0946-001</td>\n",
+ " <td>LONDON 1996-08-30</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>0</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>3</td>\n",
+ " <td>5</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>12</td>\n",
+ " <td>13</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>14</td>\n",
+ " <td>15</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " sentence_id sentence beg end\n",
+ "0 0946-001 LONDON 1996-08-30 0 1\n",
+ "1 0946-002 West Indian all-rounder Phil Simmons took four... 0 2\n",
+ "2 0946-002 West Indian all-rounder Phil Simmons took four... 3 5\n",
+ "3 0946-002 West Indian all-rounder Phil Simmons took four... 12 13\n",
+ "4 0946-002 West Indian all-rounder Phil Simmons took four... 14 15"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>sentence_id</th>\n",
+ " <th>sentence</th>\n",
+ " <th>beg</th>\n",
+ " <th>end</th>\n",
+ " <th>label</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>0946-000</td>\n",
+ " <td>CRICKET - LEICESTERSHIRE TAKE OVER AT TOP AFTE...</td>\n",
+ " <td>2</td>\n",
+ " <td>3</td>\n",
+ " <td>Leicestershire_County_Cricket_Club</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>0946-001</td>\n",
+ " <td>LONDON 1996-08-30</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>London</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>0</td>\n",
+ " <td>2</td>\n",
+ " <td>West_Indies_cricket_team</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>3</td>\n",
+ " <td>5</td>\n",
+ " <td>Phil_Simmons</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>0946-002</td>\n",
+ " <td>West Indian all-rounder Phil Simmons took four...</td>\n",
+ " <td>12</td>\n",
+ " <td>13</td>\n",
+ " <td>Leicestershire_County_Cricket_Club</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " sentence_id sentence beg end \\\n",
+ "0 0946-000 CRICKET - LEICESTERSHIRE TAKE OVER AT TOP AFTE... 2 3 \n",
+ "1 0946-001 LONDON 1996-08-30 0 1 \n",
+ "2 0946-002 West Indian all-rounder Phil Simmons took four... 0 2 \n",
+ "3 0946-002 West Indian all-rounder Phil Simmons took four... 3 5 \n",
+ "4 0946-002 West Indian all-rounder Phil Simmons took four... 12 13 \n",
+ "\n",
+ " label \n",
+ "0 Leicestershire_County_Cricket_Club \n",
+ "1 London \n",
+ "2 West_Indies_cricket_team \n",
+ "3 Phil_Simmons \n",
+ "4 Leicestershire_County_Cricket_Club "
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"df_dev_pred = df_with_pred_spans(df_dev)\n",
"display(df_dev_pred.head())"
@@ -16247,7 +16467,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 328,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -16274,8 +16494,8 @@
" quadruples consisting of the sentence id, start position, end\n",
" position and entity label of each span.\n",
" \"\"\"\n",
- " # YOUR CODE HERE\n",
- " raise NotImplementedError()"
+ " for row in df.itertuples():\n",
+ " yield row[1], row[3], row[4], row[5]"
]
},
{
@@ -16289,7 +16509,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 329,
"metadata": {
"deletable": false,
"editable": false,
@@ -16308,7 +16528,20 @@
"solution"
]
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div class=\"alert alert-success\"><strong>Checks have passed!</strong></div>"
+ ],
+ "text/plain": [
+ "<IPython.core.display.HTML object>"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"dev_gold_mentions = set(gold_mentions(df_dev))\n",
"assert ('1094-020', 0, 1, 'Seattle_Mariners') in dev_gold_mentions, \"An expected tuple is not included in the results\"\n",
@@ -16341,7 +16574,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 330,
"metadata": {
"deletable": false,
"nbgrader": {
@@ -16372,8 +16605,12 @@
" quadruples consisting of the sentence id, start position, end\n",
" position and the predicted entity label of each span.\n",
" \"\"\"\n",
- " # YOUR CODE HERE\n",
- " raise NotImplementedError()"
+ " for row in df.itertuples():\n",
+ " words = row[2].split(' ')\n",
+ " matching_wiki_name = '_'.join(words[int(row[3]):int(row[4])])\n",
+ " #sprint(matching_wiki_name)\n",
+ " yield row[1], row[3], row[4], matching_wiki_name\n",
+ " "
]
},
{
@@ -16387,7 +16624,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 331,
"metadata": {
"deletable": false,
"editable": false,
@@ -16406,7 +16643,27 @@
"solution"
]
},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Precision: 0.301, Recall: 0.274, F1: 0.287\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div class=\"alert alert-success\"><strong>Checks have passed!</strong></div>"
+ ],
+ "text/plain": [
+ "<IPython.core.display.HTML object>"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"# Compute and print the evaluation scores\n",
"scores = evaluation_scores(dev_gold_mentions, set(baseline(df_dev_pred)))\n",