Database Research Search Engine
Overview

Database Research Search Engine (beta) is a collaborative effort by people within the database research community to annotate websites related to database research. It uses Google Co-op, a general platform for annotating and querying topic-specific information.

The goal of Database Research Search Engine is to provide better search results to database research related queries that can not be effectively answered by traditional web search engines. Such queries often involve popular terms that are not specific to database research. For example, "telegraph system" (traditional results vs. topic filtered results), and "database integration" (traditional results vs. topic filtered results). By leveraging the topic-specific querying and annotation framework provided by Google Co-op, and the contributions from database researchers from around the world, we can eventually have a search engine of our own! It is also our goal to make this dataset of annotations available to the research community in the future.
Topic Context
Database Research Search Engine employs a database research context that roughly identifies important areas within the research community. The context is a small group of different facets, i.e., major aspects of the topic. Each facet contains a set of labels, which are semantic tags that users can apply to the relevant web pages. The current Database Research context is listed below. It is designed by us and undoubtedly contain our biases. We welcome comments on how to improve the context (please send your suggestions to coop-dbresearch @ <famous search engine> . com i.e., google.com).
  • People
  •   Professors
  •   Scientists
  •   Students
  • Research
  •   Groups
  •   Projects
    •   
  • Publications
  •   Papers
  •   Conferences
  •   Theses
  • Career
  •   Job Openings
  •   Job Candidates
    •   
Page Annotation
More importantly, Database Research Search Engine relies on the web page annotations. By providing a collaborative framework of easy annotation through Google Co-op and distribute the workload to each individual researchers, we hope we can, in a reasonable amount of time, collectively label most of the web pages relevant to database research. Such web pages can then be queried in a topic-sensitive way using our engine. To bootstrap the process, a small number of people have already generated a number of annotations. Some example annotations can be found here. The current set of annotations undoubtedly misses many relevant sites and even contains many erroneous annotations. As more and more researchers get involved, and start annotating sites they (and only they) know the best, the quality of the annotations will improve, so are the results.

Example Queries

People: data integration; ucb eecs
Research: telegraph; timber
Publications: data mining

How to Contribute

Contributions to Database Research Search Engine are done through Google Co-op:

  Step 1: Create a profile through the Google Co-op Profile page
  Step 2: Read Google Co-op Topic Developer Annotating Guide for introduction on the annotation process. Goto the bookmarklet page for a bookmark that allows you to annotate web pages that you are visiting. You can also submit annotations directly through the Topics page.
  Step 3: Once you have accumulated enough annotations, email your profile to coop-dbresearch@google.com. Your profile needs to be registered with Database Research Search Engine for your annotations to appear in the result; this is to prevent spam annotations corrupting the integrity of the engine.

Contact

For more information about Google Co-op, please contact Ramanathan Guha (guha @ <famous search engine> . com).

Questions, comments, and suggestions for Database Research Search Engine should be addressed to coop-dbresearch @ <famous search engine> . com, with subject line starts as "DB Research:". Alon Halevy and Cong Yu initiated this project, and created the initial topic context, along with Jayant Madhavan. Cong Yu and Arnaud Sahuguet led the annotation effort to bootstrap the process, along with Shirley Cohen, Luna Dong, and Shawn Jeffery.