SemanticEnrichment.txt

// Define your patterns
WITH [
    {label: "FinancialActivity", pattern: ["Markets", "Global stock markets"]},
    {label: "EconomicSituation", pattern: ["Credit crisis", "Mass sell-off", "Sub-prime mortgage crisis"]},
    {label: "Date", pattern: ["August 10, 2007"]},
    {label: "GeographicRegion", pattern: ["United States"]},
    {label: "FinancialIndicator", pattern: ["Dow Jones Industrial Average", "UK's FTSE-100 index", "Japan's Nikkei 225"]}
] AS patterns

// Iterate through each pattern
UNWIND patterns AS pattern

// Match nodes of type Entity
MATCH (e:Entity)

// Check if the 'id' attribute contains any of the patterns (case-insensitive)
WHERE ANY(p IN pattern.pattern WHERE toLower(e.id) CONTAINS toLower(p) OR REPLACE(toLower(e.id), '_', ' ') CONTAINS toLower(p))

// Set the corresponding label based on the pattern
SET e:DomainSpecificLabel
SET e.label = pattern.label


-------updated pattern -------------------
WITH [
    {label: "FinancialActivity", pattern: ["Markets", "Global stock markets"]},
    {label: "EconomicSituation", pattern: ["Credit crisis", "Mass sell-off", "Sub-prime mortgage crisis"]},
    {label: "Date", pattern: ["August 10, 2007"]},
    {label: "GeographicRegion", pattern: ["United States"]},
    {label: "FinancialIndicator", pattern: ["Dow Jones Industrial Average", "FTSE-100 index", "FTSE-100", "Nikkei 225", "UK's FTSE-100 index", "Japan's Nikkei 225"]}
] AS patterns

-----------------------------patterns updated on 76437----------------


// Define your patterns
WITH [
    {label: "FinancialActivity", pattern: ["Markets", "Global stock markets", "Central banks", "stockholder worries", "bail-out", "withdrawals", "investment funds", "tirade", "lower rates"]},
    {label: "EconomicSituation", pattern: ["Credit crisis", "Mass sell-off", "Sub-prime mortgage crisis", "insolvent", "volatility", "bankruptcy", "bad mortgages", "crippled", "sub-prime exposure"]},
    {label: "Date", pattern: ["August 10, 2007", "Friday", "this week"]},
    {label: "GeographicRegion", pattern: ["United States", "Europe", "Germany", "Japan", "France"]},
    {label: "FinancialIndicator", pattern: ["Dow Jones Industrial Average", "FTSE-100 index", "FTSE-100", "Nikkei 225", "U.S. Federal Reserve", "European Central Bank", "Fed", "federal funds rate", "interest rates", "share value"]},
    {label: "FinancialEntity", pattern: ["Bear Stearns", "Washington Mutual", "Countrywide Financial", "American Home Mortgage Investment Corp", "Deutsche Bank", "BNP Paribas SA"]},
    {label: "Person", pattern: ["Jim Cramer"]},
    {label: "TVNetwork", pattern: ["CNBC"]},
    {label: "TVShow", pattern: ["Mad Money", "Street Signs"]}
] AS patterns


// Iterate through each pattern
UNWIND patterns AS pattern

// Match nodes of type Entity
MATCH (e:Entity)

// Check if the 'id' attribute contains any of the patterns (case-insensitive)
WHERE ANY(p IN pattern.pattern WHERE toLower(e.id) CONTAINS toLower(p) OR REPLACE(toLower(e.id), '_', ' ') CONTAINS toLower(p))

// Set the corresponding label based on the pattern
SET e:DomainSpecificLabel
SET e.label = pattern.label
----------------------------------------------------------------------------------------------------------------

---------------------------------------updated one on 76437-----------------------------------------------------
WITH [
    {label: "FinancialActivity", patterns: ["Markets", "Global stock markets", "Central banks", "stockholder worries", "bail-out", "withdrawals", "investment funds", "tirade", "lower rates"]},
    {label: "EconomicSituation", patterns: ["Credit crisis", "Mass sell-off", "Sub-prime mortgage crisis", "subprime mortgage crisis", "insolvent", "volatility", "bankruptcy", "bad mortgages", "american home mortgage", "crippled", "sub-prime exposure"]},
    {label: "Date", patterns: ["August 10, 2007", "Friday", "this week"]},
    {label: "GeographicRegion", patterns: ["United States", "Europe", "Germany", "Japan", "France", "Asia"]},
    {label: "FinancialIndicator", patterns: ["Dow Jones Industrial Average", "FTSE-100 index", "FTSE-100", "Nikkei 225", "federal funds rate", "interest rates", "share value"]},
    {label: "FinancialEntity", patterns: ["Bear Stearns", "hedge funds", "bank of america home loans", "Washington Mutual", "Countrywide Financial", "American Home Mortgage Investment Corp", "Deutsche Bank", "BNP Paribas SA","BNP Paribas", "european central bank", "Federal Reserve"]},
    {label: "Person", patterns: ["Jim Cramer"]},
    {label: "TVNetwork", patterns: ["CNBC"]},
    {label: "TVShow", patterns: ["Mad Money", "Street Signs"]}
] AS patterns

UNWIND patterns AS pattern

MATCH (e:Entity)
WHERE ANY(p IN pattern.patterns 
          WHERE toLower(e.id) =~ toLower("(?i)\\b(" + REPLACE(p, ' ', '_') + "|" + p + ")\\b") 
            AND NOT p IN ["a", "an", "the"])

SET e:DomainSpecificLabel
SET e.label = pattern.label

----------------------------------------------------------------------------------------------------------------
-------------------------------------------------updated version of the query with market as head to be counted as FinancialActivity-----------------------

-------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------
TODO: during the entity identification phase, some exceptions need to be added in the form of rules such as U.S. 
for United States and for The U.S. and The United States. 


------------------------------UPDATED version [tested and verfied]--------------------------------------------------
// Define the mappings of labels to their corresponding entities
WITH {
    Mortgage: ["Adjustable-rate_mortgage"],
    FinancialInstitution: ["American_Home_Mortgage", "BNP_Paribas", "Bank_of_America_Home_Loans", "Bear_Stearns", "Deutsche_Bank", "Hedge_fund", "Washington_Mutual", "American Home Mortgage Investment Corp", "the Bank of Japan"],
    GeographicRegion: ["Asia", "Economy_of_Japan", "France", "Germany", "United_Kingdom", "United_States", "U.S.", "European"],
    Media: ["CNBC"],    
    CorporateEvent: ["Bankruptcy_of_Lehman_Brothers"],
    FinancialIndicator: ["Dow_Jones_Industrial_Average"],
    RegulatoryBody: ["European_Central_Bank", "Federal_Reserve", "Central banks across the world", "the government"],
    InterestRate: ["Federal_funds_rate", "interest rates", "its target rate of 5.25%", "rates"],
    Person: ["Jim_Cramer"],
    TVProgram: ["Mad_Money", "Street_Signs_(TV_program)"],
    FinancialAspect: ["Market_liquidity", "a drop in liquidity", "a greater proportion of mortgage"],
    FinancialInstrument: ["Mortgage-backed_security", "Repurchase_agreement", "shares"],
    Loan: ["Interest-only_loan", "Mortgage_loan", "loans"],
    MarketIndex: ["Nikkei_225", "The UK's FTSE-100 index"],
    Lending: ["Subprime_lending"],
    EconomicCrisis: ["Subprime_mortgage_crisis", "credit crisis", "the massive sub-prime mortgage failure"],
    MarketParticipant: ["investors"],
    MarketIndicator: ["232.90 points", "406.51 points", "more than 250 points"],
    FinancialRescue: ["a bail-out", "a record €83.6 billion addition to its banks"],
    MarketMovement: ["a mass sell-off"],
    EconomicAspect: ["a quagmire from millions of people who are unable to repay loans after taking adjustable rate mortgages, teaser rates, interest-only mortgages, or piggyback rates", "as many as seven million people"],
    RealEstateMarket: ["The U.S. housing market", "their homes"],
    TimePeriod: ["The volatile week"],
    FinancialSystem: ["Japan's financial system"],
    FinancialEvent: ["withdrawals from three large investment funds which were crippled by sub-prime exposure"],
    Market: ["Global stock markets", "Markets"]
} AS labelEntities


// Iterate through each label-entity pair and add the labels
UNWIND keys(labelEntities) AS label
unwind labelEntities[label] as entity
match (ent:Entity {id:entity}) set ent.domainType = label
//FOREACH (entity IN labelEntities[label] | 
  
//)

-------------------------------------------------------------------------------------------------------------------------------------
-----------------------------------------------updated version with more label types-------------------------------------------------
WITH{
    Mortgage: ["Adjustable-rate_mortgage"],
    FinancialInstitution: ["American_Home_Mortgage", "BNP_Paribas", "Bank_of_America_Home_Loans", "Bear_Stearns", "Deutsche_Bank", "Hedge_fund", "Washington_Mutual", "American Home Mortgage Investment Corp", "the Bank of Japan"],
    GeographicRegion: ["Asia", "Economy_of_Japan", "France", "Germany", "United_Kingdom", "United_States", "U.S.", "European"],
    Media: ["CNBC"],    
    CorporateEvent: ["Bankruptcy_of_Lehman_Brothers"],
    FinancialIndicator: ["Dow_Jones_Industrial_Average"],
    RegulatoryBody: ["European_Central_Bank", "Federal_Reserve", "Central banks across the world", "the government"],
    InterestRate: ["Federal_funds_rate", "interest rates", "its target rate of 5.25%", "rates"],
    Person: ["Jim_Cramer"],
    TVProgram: ["Mad_Money", "Street_Signs_(TV_program)"],
    FinancialAspect: ["Market_liquidity", "a drop in liquidity", "a greater proportion of mortgage"],
    FinancialInstrument: ["Mortgage-backed_security", "Repurchase_agreement", "shares"],
    Loan: ["Interest-only_loan", "Mortgage_loan", "loans"],
    MarketIndex: ["Nikkei_225", "The UK's FTSE-100 index"],
    Lending: ["Subprime_lending"],
    EconomicCrisis: ["Subprime_mortgage_crisis", "credit crisis", "the massive sub-prime mortgage failure"],
    MarketParticipant: ["investors"],
    MarketIndicator: ["232.90 points", "406.51 points", "more than 250 points"],
    FinancialRescue: ["a bail-out", "a record €83.6 billion addition to its banks"],
    MarketMovement: ["a mass sell-off"],
    EconomicAspect: ["a quagmire from millions of people who are unable to repay loans after taking adjustable rate mortgages, teaser rates, interest-only mortgages, or piggyback rates", "as many as seven million people"],
    RealEstateMarket: ["The U.S. housing market", "their homes"],
    TimePeriod: ["The volatile week"],
    FinancialSystem: ["Japan's financial system"],
    FinancialEvent: ["withdrawals from three large investment funds which were crippled by sub-prime exposure"],
    Market: ["Global stock markets", "Markets"]
} AS labelEntities 


// Iterate through each label-entity pair and add the labels
UNWIND keys(labelEntities) AS label
unwind labelEntities[label] as entity
match (ent:Entity {id:entity}) set ent.domainType = label




---------------------------*************JIm CRAmer Affiliation relation with Mad Money**************-------------------
match g= (p:Person)-[r1*2]-(:Antecedent)-[]-(:TagOccurrence {upos:'PROPN'})-[:IS_DEPENDENT {type: 'prep' }]-(:TagOccurrence {text:'of'})-[:IS_DEPENDENT {type:'pobj'}]-(e:TagOccurrence)-[:PARTICIPATES_IN]-(ne:NamedEntity )-[:REFERS_TO]-(entity:Entity)
where ne.headTokenIndex = e.tok_index_doc
merge (p)-[:AFFILIATION]-(entity)
return ne