Predicting customer churn from multi-product data
AssigneeFreshworks Inc.
InventorsSrivatsa Narasimha, Chetan Bhat, Sai Charan Rapolu, Swaminathan Padmanabhan
Life of this patent
Estimating the term…
Abstract
Predicting churn from a multi-product data using sequential modeling includes collecting, by a journey module, data from a plurality of platforms to build a journey defined by one or more events, and creating, by a sequential deep learning (DL) model, relationship data identifying between events in a chronological order, capturing temporal dependencies present in the data. This also includes correlating, by a churn prediction module, sequences of events to a churn event, and identifying when the churn event is going to occur based on pattern learnt to differentiate between “churn” and “non-churn” journeys.
Claims20
A computer-implemented method for predicting churn from a multi-product data using sequential modeling, the computer-implemented method comprising:
- training a recurrent neural network (RNN) to predict when a churn event is going to occur, wherein the training the RNN comprises
- collecting, by a journey module, historic data from a plurality of platforms to build a journey defined by one or more events,
- feeding the historic data from the plurality of platforms, wherein the historic data represents one or more events leading to a churn event,
- creating, by a sequential deep learning (DL) model, relationship data identifying between events in a chronological order, capturing temporal dependencies present in the historic data, wherein the creating the relationship data comprises
- stacking numerical data,
- transforming textual data into a numerical sequence of 0s and 1s,
- appending the stacked numerical data and the numerical sequence on a horizontal concatenation layer,
- adding usage data in a vertical concatenation layer, and
- entering horizontal concatenation and vertical concatenation as a matrix in the RNN, and
- correlating, by a churn prediction module, sequences of events from the RNN to a churn event;
- and
- predicting when the churn event is going to occur based on pattern learnt to differentiate between “churn” and “non-churn” journeys.
A system configured to predict churn from a multi-product data using sequential modeling, the system comprising:
- memory comprising a set of instructions;
- and
- at least one processor, wherein
- the set of instructions are configured to cause the at least one processor to execute
- training a recurrent neural network (RNN) to predict when a churn event is going to occur, wherein the training the RNN comprises
- collecting, by a journey module, historic data from a plurality of platforms to build a journey defined by one or more events,
- feeding the historic data from the plurality of platforms, wherein the historic data represents one or more events leading to a churn event,
- creating, by a sequential deep learning (DL) model, relationship data identifying between events in a chronological order, capturing temporal dependencies present in the historic data, wherein the creating the relationship data comprises
- stacking numerical data,
- transforming textual data into a numerical sequence of 0s and 1s,
- appending the stacked numerical data and the numerical sequence on a horizontal concatenation layer,
- adding usage data in a vertical concatenation layer, and
- entering horizontal concatenation and vertical concatenation as a matrix in the RNN, and
- correlating, by a churn prediction module, sequences of events from the RNN to a churn event;
- and
- predicting when the churn event is going to occur based on pattern learnt to differentiate between “churn” and “non-churn” journeys.
A non-transitory computer-readable medium comprising a computer program that predicts churn from a multi-product data using sequential modeling, the computer program is configured to cause at least one processor to execute:
- training a recurrent neural network (RNN) to predict when a churn event is going to occur, wherein the training the RNN comprises
- collecting, by a journey module, historic data from a plurality of platforms to build a journey defined by one or more events,
- feeding the historic data from the plurality of platforms, wherein the historic data represents one or more events leading to a churn event,
- creating, by a sequential deep learning (DL) model, relationship data identifying between events in a chronological order, capturing temporal dependencies present in the historic data, wherein the creating the relationship data comprises
- stacking numerical data,
- transforming textual data into a numerical sequence of 0s and 1s,
- appending the stacked numerical data and the numerical sequence on a horizontal concatenation layer,
- adding usage data in a vertical concatenation layer, and
- entering horizontal concatenation and vertical concatenation as a matrix in the RNN, and
- correlating, by a churn prediction module, sequences of events from the RNN to a churn event, and
- predicting when the churn event is going to occur based on pattern learnt to differentiate between “churn” and “non-churn” journeys.
Dependencies are inferred from the claim text.