Alexandrin Popescul

Methods and metrics for cold-start recommendations

We have developed a method for recommending items that combines content and collaborative data under a single probabilistic framework. We benchmark our algorithm against a naïve Bayes classifier on the cold-start problem, where we wish to recommend items that no one in the community has yet rated. We systematically explore three testing methodologies using a publicly available data set, and explain how these methods apply to specific real-world applications. We advocate heuristic recommenders when benchmarking to give competent baseline performance. We introduce a new performance metric, the CROC curve, and demonstrate empirically that the various components of our testing strategy combine to obtain deeper understanding of the performance characteristics of recommender systems. Though the emphasis of our testing is on cold-start recommending, our methods for recommending and evaluation are general.

Clustering and identifying temporal trends in document databases

We introduce a simple and efficient method for clustering and identifying temporal trends in hyper-linked document databases. Our method can scale to large datasets because it exploits the underlying regularity often found in hyper-linked document databases. Because of this scalability, we can use our method to study the temporal trends of individual clusters in a statistically meaningful manner. As an example of our approach, we give a summary of the temporal trends found in a scientific literature database with thousands of documents.

CROC: A New Evaluation Criterion for Recommender Systems

Evaluation of a recommender system algorithm is a challenging task due to the many possible scenarios in which such systems may be deployed. We have designed a new performance plot called the CROC curve with an associated statistic: the area under the curve. Our CROC curve supplements the widely used ROC curve in recommender system evaluation by discovering performance characteristics that standard ROC evaluation often ignores. Empirical studies on two domains and including several recommender system algorithms demonstrate that combining ROC and CROC curves in evaluation can lead to a more informed characterization of performance than using either curve alone.

Parallel Corpus Approach for Name Matching in Record Linkage

Record linkage, or entity resolution, is an important area of data mining. Name matching is a key component of systems for record linkage. Alternative spellings of the same name are a common occurrence in many applications. We use the largest collection of genealogy person records in the world together with user search query logs to build name-matching models. The procedure for building a crowd-sourced training set is outlined together with the presentation of our method. We cast the problem of learning alternative spellings as a machine translation problem at the character level. We use information retrieval evaluation methodology to show that this method substantially outperforms on our data a number of standard well known phonetic and string similarity methods in terms of precision and recall. Our result can lead to a significant practical impact in entity resolution applications.