System and method for classifying documents

A non-transitory computer readable medium having computer executable program code embodied thereon, the computer executable program code configured to cause a computing device to perform the operations of:

• partitioning a digital real estate document into a set of token patterns;
• normalizing the set of token patterns to obtain a normalized set of token patterns;
• applying a first filter to the normalized set of token patterns to obtain a first indicator of whether the digital real estate document is a member of a first document classification and a first confidence score for the first indicator;
• applying a second filter to the normalized set of token patterns to obtain a second indicator of whether the digital real estate document is a member of a second document classification and a second confidence score for the second indicator;
• and
• determining a document classification of the digital real estate document from a set of indicators associated with the digital real estate document and set of confidence scores associated with the set of indicators, wherein the set of indicators includes the first and second indicators and the set of confidence scores includes the first and second confidence scores;
• wherein the first filter comprises a set of indicative token patterns and a set of non-indicative token patterns, wherein the indicative token patterns are those associated with the first document classification, wherein the non-indicative token patterns are those not associated with the first document classification;
• and
• wherein applying the first filter to the normalized set of token patterns to obtain the first indicator and the first confidence score comprises:
• performing a determination of whether the digital real estate document is a member of the first document classification based on the set of indicative token patterns, the set of non-indicative token patterns, and the normalized set of token patterns;
• setting the first indicator according to the determination;
• and
• calculating the first confidence score based on a level of confidence in the determination.

A computer system configured to determine a document classification for a digital real estate document, comprising:

• a processor;
• a memory connected to the processor;
• and
• a computer-readable medium having instructions embedded therein, instructions configured to cause the processor to perform the operations of:
• partitioning the digital real estate document into a set of token patterns;
• normalizing the set of token patterns to obtain a set of normalized set of token patterns;
• applying a first filter to the normalized set of token patterns to obtain a first score for the digital real estate document, wherein the first score indicates a first likelihood that the digital real estate document is a member of a first document classification associated with the first filter;
• applying a second filter to the normalized set of token patterns to obtain a second score for the digital real estate document, wherein the second score indicates a second likelihood that the digital real estate document is a member of a second document classification associated with the second filter;
• and
• determining the document classification of the digital real estate document based on a set of scores associated with the digital real estate document, wherein the set of scores includes the first score and the second score;
• and
• wherein the first filter comprises a set of indicative token patterns and a set of non-indicative token patterns, wherein the indicative token patterns are those associated with the first document classification, wherein the non-indicative token patterns are those not associated with the first document classification, and wherein applying the first filter to the normalized set of token patterns to obtain the first score comprises determining the first score based on the set of indicative token patterns, the set of non-indicative token patterns, and the normalized set of token patterns.

A computer system configured to determine a document classification for a digital real estate document, comprising:

• a processor;
• a memory connected to the processor;
• and
• a computer-readable medium having instructions embedded therein, instructions configured to cause the processor to perform the operations of:
• partitioning the digital real estate document into a set of token patterns;
• normalizing the set of token patterns to obtain a set-ef normalized set of token patterns;
• applying a first filter to the normalized set of token patterns to obtain a first indicator of whether the digital real estate document is a member of a first document classification and a first confidence score for the first indicator;
• applying a second filter to the normalized set of token patterns to obtain a second indicator of whether the digital real estate document is a member of a second document classification and a second confidence score for the first Indicator;
• and
• determining the document classification of the digital real estate document from a set of indicators associated with the digital real estate document and set of confidence scores associated with the set of indicators, wherein the set of indicators includes the first and second indicators and the set of confidence scores includes the first and second confidence scores;
• wherein the first filter comprises a set of indicative token patterns and a set of non-indicative token patterns, wherein the indicative token patterns are those associated with the first document classification, wherein the non-indicative token patterns are those not associated with the first document classification.