MATCHING VISITORS AS LEADS TO LEAD BUYERS
An example system comprises a memory that stores instructions, a graphical user interface, a matcher simulator, a matching engine, and a set of server computers storing instructions that cause the set of server computers to perform certain operations. The operations include accessing, at the matcher simulator, a dataset comprising observed and unobserved variables, the observed variables including visitor profiles the unobserved variables including visitor responses to presented filters having filter criteria, dividing the dataset into a training set and a value set, training a discriminative model or a generative model using the training set, using the trained discriminative or generative model, running the matcher simulator against various instantiations of the matching engine and capturing impacts of changes in the filter criteria and based on at least one captured impact determining a placement position of a splitting question in a form presented to a user in the graphical user interface.
1 . A system comprising:
a memory that stores instructions;
a graphical user interface;
a matcher simulator;
a matching engine; and
a set of server computers storing instructions that, when executed by one or more processors of the set of server computers, causes the set of server computers to perform operations comprising, at least:
accessing, at the matcher simulator, a dataset comprising observed and unobserved variables, the observed variables including visitor profiles, the unobserved variables including visitor responses to presented filters having filter criteria;
dividing the dataset into a training set and a value set;
training a discriminative model or a generative model using the training set;
using the trained discriminative or generative model, running the matcher simulator against various instantiations of the matching engine and capturing impacts of changes in the filter criteria; and
based on at least one captured impact, determining a placement position of a splitting question in a form presented to a user in the graphical user interface.
2 . The system of claim 1 , wherein the matcher simulator is configured to simulate an operation of the system over a given time frame.
3 . The system of claim 2 , wherein the matcher simulator is configured to run previously collected visitor data through the matching engine.
4 . The system of claim 3 , wherein the operations further comprise estimating parameters of a distribution of revenue based on the previously collected visitor data.
5 . The system of claim 1 , wherein the operations further comprise presenting alternate instances of the form containing the splitting question to the user, or other users.
6 . The system of claim 1 , wherein the placement position of the splitting question on the form is based further on a quality of user traffic.
7 . A method comprising:
accessing, at a matcher simulator, a dataset comprising observed and unobserved variables, the observed variables including visitor profiles, the unobserved variables including visitor responses to presented filters having filter criteria;
dividing the dataset into a training set and a value set;
training a discriminative model or a generative model using the training set;
using the trained discriminative or generative model, running the matcher simulator against various instantiations of a matching engine and capturing impacts of changes in the filter criteria; and
based on at least one captured impact, determining a placement position of a splitting question in a form presented to a user in a graphical user interface.
8 . The method of claim 7 , further comprising simulating an operation of the system over a given time frame.
9 . The method of claim 8 , further comprising running previously collected visitor data through the matching engine.
10 . The method of claim 9 , further comprising estimating parameters of a distribution of revenue based on the previously collected visitor data.
11 . The method of claim 7 , further comprising presenting alternate instances of the form containing the splitting question to the user, or other users.
12 . The method of claim 7 , wherein the placement position of the splitting question on the form is based further on a quality of user traffic.
13 . A machine-readable medium including instructions which, when read by a machine, cause the machine to perform operations comprising:
accessing, at a matcher simulator, a dataset comprising observed and unobserved variables, the observed variables including visitor profiles, the unobserved variables including visitor responses to presented filters having filter criteria;
dividing the dataset into a training set and a value set;
training a discriminative model or a generative model using the training set;
using the trained discriminative or generative model, running the matcher simulator against various instantiations of a matching engine and capturing impacts of changes in the filter criteria; and
based on at least one captured impact, determining a placement position of a splitting question in a form presented to a user in a graphical user interface.