Last year we added query suggestions to help students explore topics they may not be familiar with. These suggestions go from a broad search to deeper and more specific areas. But many of you are already well familiar with your research area, and your searches are already specific and detailed. Sometimes, it's good to take a step back and go into a different, but related, space.

Today, we're adding query suggestions for detailed queries. They help researchers explore topics related to the original query. For example, consider the query suggestions for [semantic segmentation object detection]. They cover:
Semantic segmentation: [semantic segmentation rgb d images], [fully convolutional networks for semantic segmentation], [deep structured models for semantic segmentation], [indoor semantic segmentation], [fast semantic segmentation], [semantic segmentation scene classification], [semantic segmentation deconvolution network]
Object detection: [localization accuracy object detection], [joint object detection], [real time object detection]
Combination of concepts: [rich features object detection and segmentation], [semantic segmentation context for object detection]

Note that query suggestions appear below search results.
The new query suggestions span all broad areas of research. For example, see [prions protein folding], [global stock market portfolio selection], [test salmonella spp], [racial discrimination and gerrymandering], [gamma irradiation diamond detector], [binary planet formation], [aspect based sentiment analysis], [axial flow turbojet engine].

For now, the additional suggestions are limited to English queries. We plan to expand the coverage to more languages.

Source: Google Scholar Blog