Employer name normalization, or linking employer names in job postings or resumes to entities in an employer knowledge base (KB), is important for many downstream applications in the online recruitment domain. Key challenges for employer name normalization include handling employer names from both job postings and resumes, leveraging the corresponding location and URL context, and handling name variations and duplicates in the KB. In this paper, we describe the CompanyDepot system developed at CareerBuilder, which uses machine learning techniques to address these challenges. We discuss the main challenges and share our lessons learned in deployment, maintenance, and utilization of the system over the past two years. We also share several examples of how the system has been used in applications at CareerBuilder to deliver value to end customers.

We describe a method for automatically generating subjectivity clues for a specific topic and a set of (relevant) document, evaluating it on the task of classifying sentences w.r.t. subjectivity, with improvements over previous work.

We present a corpus-based approach to the class expansion task. For a given set of seed entities we use co-occurrence statistics taken from a text collection to define a membership function that is used to rank candidate entities for inclusion in the set. We describe an evaluation framework that uses data from Wikipedia. The performance of our class extension method improves as the size of the text collection increases.

We address the task of extracting focused salient information items, relevant and important for a given topic, from a large encyclopedic resource. Specifically, for a given topic (a Wikipedia article) we identify snippets from other articles in Wikipedia that contain important information for the topic of the original article, without duplicates. We compare several methods for addressing the task, and find that a mixture of content-based, link-based, and layout-based features outperforms other methods, especially in combination with the use of so-called reference corpora that capture the key properties of entities of a common type.

The Netherlands had parliamentary elections on November 22, 2006. We built a system which helped voters to make an informed choice among the many participating parties. One of the most important pieces of information in the Dutch election and subsequent coalition government formation is the party program, a text document with an average length of 45 pages. Our system provides the voter with focused access to party programs, enabling her to make a topic-wise comparison of parties' viewpoints. We complemented this type of access ("What do the parties promise?") with access to news ("What happens around these topics?") and blogs ("What do people say about them?"). We describe the system, including design technical details, and user statistics.

We address the task of answering natural language questions by using the large number of Frequently Asked Questions (FAQ) pages available on the web. The task involves three steps: (1) fetching FAQ pages from the web; (2) automatic extraction of question/answer (Q/A) pairs from the collected pages; and (3) answering users' questions by retrieving appropriate Q/A pairs. We discuss our solutions for each of the three tasks, and give detailed evaluation results on a collected corpus of about 3.6Gb of text data (293K pages, 2.8M Q/A pairs), with real users' questions sampled from a web search engine log. Specifically, we propose simple but effective methods for Q/A extraction and investigate task-specific retrieval models for answering questions. Our best model finds answers for 36% of the test questions in the top 20 results. Our overall conclusion is that FAQ pages on the web provide an excellent resource for addressing real users' information needs in a highly focused manner.

We describe a method for enriching the output of a parser with information available in a corpus. The method is based on graph rewriting using memory-based learning, applied to dependency structures. This general framework allows us to accurately recover both grammatical and semantic information as well as non-local dependencies. It also facilitates dependency-based evaluation of phrase structure parsers. Our method is largely independent of the choice of parser and corpus, and shows state of the art performance.