Efficient query recommendations in the long tail via center-piece subgraphs

Abstract

We present a recommendation method based on the wellknown concept of center-piece subgraph, that allows for the time/space efficient generation of suggestions also for rare, i.e., long-tail queries. Our method is scalable with respect to both the size of datasets from which the model is computed and the heavy workloads that current web search engines have to deal with. Basically, we relate terms contained into queries with highly correlated queries in a query-flow graph. This enables a novel recommendation generation method able to produce recommendations for approximately 99% of the workload of a real-world search engine. The method is based on a graph having term nodes, query nodes, and two kinds of connections: term-query and query-query. The first connects a term to the queries in which it is contained, the second connects two query nodes if the likelihood that a user submits the second query after having issued the first one is sufficiently high. On such large graph we need to compute the center-piece subgraph induced by terms contained into queries. In order to reduce the cost of the above computation, we introduce a novel and efficient method based on an inverted index representation of the model. We experiment our solution on two real-world query logs and we show that its effectiveness is comparable (and in some case better) than state-of-the-art methods for head-queries. More importantly, the quality of the recommendations generated remains very high also for long-tail queries, where other methods fail even to produce any suggestion. Finally, we extensively investigate scalability and efficiency issues and we show the viability of our method in real world search engines.

Task-aware query recommendation

“…To increase the speed of our recommendation, we only stored the 100,000 top scoring random walk results for each term. Bonchi et al [3] found this to have no or very limited effects on performance when used with c = 0.1.…”

“…We only consider one query recommendation algorithm based largely on the query flow graph (QFG) work of Boldi et al [2] and the term-query graph (TQGraph) work of Bonchi et al [3]. We use a query flow graph G in which the vertices V are queries from a query log L and the edges E represent reformulation probabilities.…”

“…To generate recommendations, we rely on a slight adaptation of the query flow graph, called the term-query graph [3]. This adds a layer to the QFG that consists of all terms that occur in queries in the QFG, each of which points to the queries in which it occurs.…”

“…For the query recommendation using the TQGraph, we used a restart probability of c = 0.1, as was found to be optimal by Bonchi et al [3]. Note that they refer to the restart value α, where c = 1 − α.…”

“…analysis of task-aware query recommendation demonstrating the usefulness of on-task query context, (2) an analysis of the impact of automatic search task identification on taskaware recommendation, in which we show the state of the art works very well, (3) an open source data set for evaluating context-aware query recommendation, and (4) a generalized model of combining recommendations across a search context, regardless of the recommendation algorithm.…”