|
151 | 151 | "# Define document categories and their descriptions\n",
|
152 | 152 | "classifier_schema = {\n",
|
153 | 153 | " \"categories\": {\n",
|
154 |
| - " \"Loan application\": {\n", |
| 154 | + " \"Loan application\": { # Both space and underscore are allowed in category names\n", |
155 | 155 | " \"description\": \"Documents submitted by individuals or businesses to request funding, typically including personal or business details, financial history, loan amount, purpose, and supporting documentation.\"\n",
|
156 | 156 | " },\n",
|
157 | 157 | " \"Invoice\": {\n",
|
158 | 158 | " \"description\": \"Billing documents issued by sellers or service providers to request payment for goods or services, detailing items, prices, taxes, totals, and payment terms.\"\n",
|
159 | 159 | " },\n",
|
160 |
| - " \"Bank_Statement\": {\n", |
| 160 | + " \"Bank_Statement\": { # Both space and underscore are allowed in category names\n", |
161 | 161 | " \"description\": \"Official statements issued by banks that summarize account activity over a period, including deposits, withdrawals, fees, and balances.\"\n",
|
162 | 162 | " },\n",
|
163 | 163 | " },\n",
|
|
347 | 347 | " \"enableLayout\": True, # Extract layout information\n",
|
348 | 348 | " \"enableBarcode\": False, # Skip barcode detection\n",
|
349 | 349 | " \"enableFormula\": False, # Skip formula detection\n",
|
350 |
| - " \"estimateFieldSourceAndConfidence\": False, # Set to True if you want to estimate the field location (aka grounding) and confidence\n", |
| 350 | + " \"estimateFieldSourceAndConfidence\": True, # Set to True if you want to estimate the field location (aka grounding) and confidence\n", |
351 | 351 | " \"disableContentFiltering\": False,\n",
|
352 | 352 | " },\n",
|
353 | 353 | " \"fieldSchema\": {\n",
|
|
429 | 429 | "# Create enhanced classifier schema\n",
|
430 | 430 | "enhanced_classifier_schema = {\n",
|
431 | 431 | " \"categories\": {\n",
|
432 |
| - " \"Loan application\": {\n", |
| 432 | + " \"Loan application\": { # Both space and underscore are allowed in category names\n", |
433 | 433 | " \"description\": \"Documents submitted by individuals or businesses to request funding, typically including personal or business details, financial history, loan amount, purpose, and supporting documentation.\",\n",
|
434 | 434 | " \"analyzerId\": analyzer_id # IMPORTANT: Use created custom analyzer in previous step for load applications\n",
|
435 | 435 | " },\n",
|
436 | 436 | " \"Invoice\": {\n",
|
437 | 437 | " \"description\": \"Billing documents issued by sellers or service providers to request payment for goods or services, detailing items, prices, taxes, totals, and payment terms.\",\n",
|
438 | 438 | " \"analyzerId\": \"prebuilt-invoice\" # IMPORTANT: Use created custom analyzer in previous step for annual reports\n",
|
439 | 439 | " },\n",
|
440 |
| - " \"Bank_Statement\": {\n", |
| 440 | + " \"Bank_Statement\": { # Both space and underscore are allowed in category names\n", |
441 | 441 | " \"description\": \"Official statements issued by banks that summarize account activity over a period, including deposits, withdrawals, fees, and balances.\"\n",
|
442 | 442 | " # No analyzer specified - uses default processing\n",
|
443 | 443 | " }\n",
|
|
0 commit comments