Automatic detection of fraudulent real estate listings
A system for rental fraud detection comprises an input interface, a feature vector extractor, a model builder, and a predictor. The input interface is to receive a set of training data, wherein the set of training data comprises a set of rental fraud indications and to receive a set of prediction data. The feature vector extractor is to extract a first training set of feature vectors from the set of training data and to extract a first test feature vector from the set of prediction data. The model builder is to build a first model using the first training set of feature vectors and the rental fraud indications. The predictor is to predict whether the first test feature vector corresponds to fraudulent data using the first model.
1. A system for rental fraud detection, comprising:
an input interface configured to:
receive a set of training data, wherein the set of training data comprises a set of rental fraud indications; and
receive a set of prediction data;
a feature vector extractor configured to:
extract a first training set of feature vectors from the set of training data, wherein the set of training data comprises real estate listings determined to be not fraudulent and real estate listings determined to be fraudulent; and
extract a first test feature vector from the set of prediction data, wherein the set of prediction data comprises one or more real estate listings of undetermined fraudulence;
a model builder configured to:
build a first model using the first training set of feature vectors and the rental fraud indications; and
a predictor configured to:
predict whether the first test feature vector corresponds to fraudulent data using the first model.
2. The system of claim 1 , wherein the feature vector extractor is further configured to:
extract a second training set of feature vectors from the set of training data; and
extract a second test feature vector from the set of prediction data.
3. The system of claim 2 , wherein the model builder is further configured to:
build a second model using the second training set of feature vectors and the rental fraud indications.
4. The system of claim 3 , wherein the predictor is further configured to:
predict whether the first test feature vector and the second test feature vector correspond to fraudulent data using the first model and the second model.
5. The system of claim 3 , wherein the predictor is further configured to:
predict whether the first test feature vector and the second test feature vector correspond to fraudulent data using a weighted combination of the first model and the second model.
6. The system of claim 5 , wherein the weight associated with the first model comprises 0.25 and the weight associated with the second model comprises 0.75.
7. The system of claim 1 , wherein a feature of the first test feature vector comprises a property attribute.
8. The system of claim 7 , wherein the property attribute comprises a property type.
9. The system of claim 7 , wherein the property attribute comprises a number of bedrooms.
10. The system of claim 1 , wherein a feature of the first test feature vector comprises a property location attribute.
11. The system of claim 10 , wherein the property location attribute comprises a zip code.
12. The system of claim 10 , wherein the property location attribute comprises a three digit zip code prefix.
13. The system of claim 1 , wherein a feature of the first test feature vector comprises a property price attribute.
14. The system of claim 13 , wherein the property price attribute comprises a property price percentile.
15. The system of claim 1 , wherein a feature of the first test feature vector comprises a property agent attribute.
16. The system of claim 1 , wherein a feature of the first test feature vector comprises a fraud rate.
17. The system of claim 1 , wherein a feature of the first test feature vector comprises a property lister attribute.
18. The system of claim 2 , wherein the second test feature vector and the first test feature vector comprise distinct features.
19. A method, performed by a computing system having a memory and a processor, for rental fraud detection, comprising:
receiving, with the processor, a set of training data, wherein the set of training data comprises a set of rental fraud indications;
receiving, with the processor, a set of prediction data;
extracting, with the processor, a first training set of feature vectors from the set of training data, wherein the set of training data comprises real estate listings determined to be not fraudulent and real estate listings determined to be fraudulent;
extracting, with the processor, a first test feature vector from the set of prediction data, wherein the set of prediction data comprises one or more real estate listings each of undetermined fraudulence;
building, with the processor, a first model using the first training set of feature vectors and the rental fraud indications; and
predicting, with the processor, whether the first test feature vector corresponds to fraudulent data using the first model.
20. A computer program product for rental fraud detection, the computer program product being embodied in a computer readable storage medium and comprising computer instructions for:
receiving a set of training data, wherein the set of training data comprises a set of rental fraud indications;
receiving a set of prediction data;
extracting a first training set of feature vectors from the set of training data, wherein the set of training data comprises real estate listings determined to be not fraudulent and real estate listings determined to be fraudulent;
extracting a first test feature vector from the set of prediction data, wherein the set of prediction data comprises one or more real estate listings each of undetermined fraudulence;
building a first model using the first training set of feature vectors and the rental fraud indications; and
predicting whether the first test feature vector corresponds to fraudulent data using the first model.