Description
IMPORTANT: Don’t run this model on the whole financial report. Instead:
- Split by paragraphs;
- Use the
finclf_form_10k_summary_itemText Classifier to select only these paragraps;
This Financial NER Model is aimed to process the first summary page of 10K filings and extract the information about the Company submitting the filing, trading data, address / phones, CFN, IRS, etc.
This is another version of finner_10k_summary trained with better financial embeddings (bert_embeddings_sec_bert_base)
Predicted Entities
ADDRESS, CFN, FISCAL_YEAR, IRS, ORG, PHONE, STATE, STOCK_EXCHANGE, TICKER, TITLE_CLASS, TITLE_CLASS_VALUE
How to use
document_assembler = nlp.DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
sentence_detector = nlp.SentenceDetector() \
.setInputCols(["document"]) \
.setOutputCol("sentence") \
.setCustomBounds(["\n\n"])
tokenizer = nlp.Tokenizer()\
.setInputCols(["sentence"])\
.setOutputCol("token")
embeddings = nlp.BertEmbeddings.pretrained("bert_embeddings_sec_bert_base","en")\
.setInputCols(["sentence", "token"])\
.setOutputCol("embeddings")\
.setCaseSensitive(True)\
.setMaxSentenceLength(512)
ner_model = finance.NerModel.pretrained("finner_sec_10k_summary","en","finance/models")\
.setInputCols(["sentence", "token", "embeddings"])\
.setOutputCol("ner")\
ner_converter = nlp.NerConverter()\
.setInputCols(["sentence", "token", "ner"])\
.setOutputCol("ner_chunk")
pipeline = nlp.Pipeline(stages=[
document_assembler,
sentence_detector,
tokenizer,
embeddings,
ner_model,
ner_converter
])
model = pipeline.fit(spark.createDataFrame([[""]]).toDF("text"))
data = spark.createDataFrame([["""ANNUAL REPORT PURSUANT TO SECTION 13 OR 15(d) OF THE SECURITIES AND EXCHANGE ACT OF 1934
For the annual period ended January 31, 2021
or
TRANSITION REPORT PURSUANT TO SECTION 13 OR 15(d) OF THE SECURITIES EXCHANGE ACT OF 1934
For the transition period from________to_______
Commission File Number: 001-38856
PAGERDUTY, INC.
(Exact name of registrant as specified in its charter)
Delaware
27-2793871
(State or other jurisdiction of
incorporation or organization)
(I.R.S. Employer
Identification Number)
600 Townsend St., Suite 200, San Francisco, CA 94103
(844) 800-3889
(Address, including zip code, and telephone number, including area code, of registrant’s principal executive offices)
Securities registered pursuant to Section 12(b) of the Act:
Title of each class
Trading symbol(s)
Name of each exchange on which registered
Common Stock, $0.000005 par value,
PD
New York Stock Exchange"""]]).toDF("text")
result = model.transform(data)
result.select(F.explode(F.arrays_zip('ner_chunk.result', 'ner_chunk.metadata')).alias("cols")) \
.select(F.expr("cols['0']").alias("ticker"),
F.expr("cols['1']['entity']").alias("label")).show(50, truncate = False)
Results
+----------------------------------------------+-----------------+
|ticker |label |
+----------------------------------------------+-----------------+
|January 31, 2021 |FISCAL_YEAR |
|001-38856 |CFN |
|PAGERDUTY, INC |ORG |
|Delaware |STATE |
|27-2793871 |IRS |
|600 Townsend St., Suite 200, San Francisco, CA|ADDRESS |
|(844) 800-3889 |PHONE |
|Common Stock |TITLE_CLASS |
|$0.000005 |TITLE_CLASS_VALUE|
|PD |TICKER |
|New York Stock Exchange |STOCK_EXCHANGE |
+----------------------------------------------+-----------------+
Model Information
| Model Name: | finner_sec_10k_summary |
| Compatibility: | Finance NLP 1.0.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, embeddings] |
| Output Labels: | [ner] |
| Language: | en |
| Size: | 16.5 MB |
References
Manual annotations on 10-K Filings
Benchmarking
label precision recall f1-score support
B-TITLE_CLASS 1.00 1.00 1.00 15
I-TITLE_CLASS 1.00 1.00 1.00 21
B-ORG 0.84 0.66 0.74 62
I-ORG 0.88 0.76 0.82 93
B-STOCK_EXCHANGE 0.86 0.86 0.86 14
I-STOCK_EXCHANGE 0.98 0.98 0.98 50
B-PHONE 0.95 0.87 0.91 23
I-PHONE 0.95 1.00 0.98 60
B-STATE 0.89 0.85 0.87 20
B-IRS 1.00 0.88 0.93 16
B-ADDRESS 0.94 0.83 0.88 18
I-ADDRESS 0.92 0.97 0.94 144
B-TICKER 0.86 0.92 0.89 13
B-FISCAL_YEAR 0.96 0.88 0.92 50
I-FISCAL_YEAR 0.93 0.92 0.92 125
B-TITLE_CLASS_VALUE 1.00 0.93 0.97 15
B-CFN 0.92 1.00 0.96 12
micro-avg 0.93 0.89 0.91 751
macro-avg 0.84 0.81 0.82 751
weighted-avg 0.92 0.89 0.91 751