Title / Abstract Extraction Tool

Langchain tools that execute zero-shot extraction tasks over a local database of title/abstracts from papers previously imported into our database.

source

BaseTitleAbstractExtractionTool

 BaseTitleAbstractExtractionTool
                                  (db:alhazen.utils.ceifns_db.Ceifns_Liter
                                  atureDb, llm:Optional[langchain_core.lan
                                  guage_models.chat_models.BaseChatModel]=
                                  None, slm:Optional[langchain_core.langua
                                  ge_models.chat_models.BaseChatModel]=Non
                                  e, name:str='tiab_classification',
                                  description:str='Runs a specified
                                  document classification pipeline over
                                  papers in a collection.', args_schema:Op
                                  tional[Type[pydantic.v1.main.BaseModel]]
                                  =None, return_direct:bool=True,
                                  verbose:bool=False, callbacks:Union[List
                                  [langchain_core.callbacks.base.BaseCallb
                                  ackHandler],langchain_core.callbacks.bas
                                  e.BaseCallbackManager,NoneType]=None, ca
                                  llback_manager:Optional[langchain_core.c
                                  allbacks.base.BaseCallbackManager]=None,
                                  tags:Optional[List[str]]=None,
                                  metadata:Optional[Dict[str,Any]]=None, h
                                  andle_tool_error:Union[bool,str,Callable
                                  [[langchain_core.tools.ToolException],st
                                  r],NoneType]=False, handle_validation_er
                                  ror:Union[bool,str,Callable[[pydantic.v1
                                  .error_wrappers.ValidationError],str],No
                                  neType]=False, prompt_name:str='binary
                                  methods paper', examples:dict={})

Runs a specified document classification pipeline over papers in a collection.


source

TitleAbstractExtractionToolSchema

 TitleAbstractExtractionToolSchema (collection_id:str, domain:str,
                                    extraction_type:str,
                                    repeat_run:Optional[bool]=None)

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.


source

TitleAbstractExtraction_OneDocAtATime_Tool

 TitleAbstractExtraction_OneDocAtATime_Tool
                                             (db:alhazen.utils.ceifns_db.C
                                             eifns_LiteratureDb, llm:Optio
                                             nal[langchain_core.language_m
                                             odels.chat_models.BaseChatMod
                                             el]=None, slm:Optional[langch
                                             ain_core.language_models.chat
                                             _models.BaseChatModel]=None, 
                                             name:str='tiab_one_doc_classi
                                             fication',
                                             description:str='Runs a
                                             specified document
                                             classification pipeline over
                                             papers in a collection by
                                             running a simple classifier
                                             over the text of each title +
                                             abstract.', args_schema:Optio
                                             nal[Type[pydantic.v1.main.Bas
                                             eModel]]=None,
                                             return_direct:bool=True,
                                             verbose:bool=False, callbacks
                                             :Union[List[langchain_core.ca
                                             llbacks.base.BaseCallbackHand
                                             ler],langchain_core.callbacks
                                             .base.BaseCallbackManager,Non
                                             eType]=None, callback_manager
                                             :Optional[langchain_core.call
                                             backs.base.BaseCallbackManage
                                             r]=None, tags:Optional[List[s
                                             tr]]=None, metadata:Optional[
                                             Dict[str,Any]]=None, handle_t
                                             ool_error:Union[bool,str,Call
                                             able[[langchain_core.tools.To
                                             olException],str],NoneType]=F
                                             alse, handle_validation_error
                                             :Union[bool,str,Callable[[pyd
                                             antic.v1.error_wrappers.Valid
                                             ationError],str],NoneType]=Fa
                                             lse, prompt_name:str='binary
                                             methods paper',
                                             examples:dict={})

Runs a specified document classification pipeline over papers in a collection.