On Correcting Misspelled Queries in Email Search

Authors: Abhijit Bhole, Raghavendra Udupa

AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We do a systematic empirical study of Sp EQ on two public domain email corpora. We compare the performance of Spe Q against two baselines a source-channel based ranking approach and the spelling correction feature of a popular email service. Our experiments show that Sp EQ gives significant improvement in accuracy over the two baselines. We did a feature ablative study and found that each class of feature functions contributed to the effectiveness of our approach.
Researcher Affiliation Industry Abhijit Bhole and Raghavendra Udupa Microsoft Research 9 Lavelle Road, Bangalore 560001 {a-abhina, raghavu}@microsoft.com
Pseudocode No The paper describes the components of its ranking framework and feature functions but does not include a pseudocode block or algorithm.
Open Source Code No The paper does not provide any information about the availability of its source code.
Open Datasets Yes Data We used two public domain email corpora for our experiments: PALIN: About 15, 000 mails of the former Alaskan governor Sarah Palin (from the period December 2006 to September 2008 made public in the year 2011) in two folders Inbox and Sent Mail. ENRON: All mails (about 2100) in the All Documents folder of rogers-b, an employee of Enron corporation.
Dataset Splits No The paper mentions training data but does not specify any explicit training/validation/test splits, percentages, or absolute counts for reproduction.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers.
Experiment Setup No The paper describes the components of its Sp EQ framework and feature functions but does not provide specific experimental setup details such as hyperparameters (e.g., learning rate, batch size, number of epochs) or other system-level training settings.