Yahoo! Inc.

Jollify Management Limited

Excalibur Ip, LLC

Richard Kasperski

Ralph Rabbat

Shyam Kapur

Ashley Woodman Hall

Mihir Shah

Oliver Thomas Bayley

John Thrall

Alastair Robert Gourlay

Peter Anick

Stephen L Hood

Adam Durfee

Benoit F Dumoulin

Farzin Maghoul

Omer Emre Velipasaoglu

Su-Lin Wu

Arkady Borkovsky

Jiefeng Shen

Chi-Hoon Lee

Eric Crestan

Providing a speculative search result for a search query prior to completion of the search query is described. In response to receiving a search query from a client node, a speculative search result is provided to the client node for the search query prior to receiving an indication from the client node that said search query is completely formed. The speculative search result may be displayed on the same web page on the client node as the search query, while the search query is being entered by the user. As the user further enters the search query, a new speculative search result may be provided to the user.

Speculative search result on a not-yet-submitted search query

Techniques are described herein for providing locality-sensitive search suggestions. When users issue search queries, the localities of the users who issue those search queries are associated with the search queries. The users may be grouped into populations, such that each population has a locality that corresponds to the localities of the users in that population. The issuances of the search queries induce a distribution of interest with respect to the issued search queries among the populations. A subset (e.g., one or more) of the issued search queries is suggested to a user who provides a search query (e.g., a preliminary search query or an issued search query), which is deemed to be relevant to the subset of the issued search queries, based on the distribution of interest.

Locality-sensitive search suggestions

Providing a speculative search result for a search query prior to completion of the search query. In response to receiving a search query from a client node, a speculative search result is provided to the client node for the search query prior to receiving an indication from the client node that said search query is completely formed. The speculative search result may be displayed on the same web page on the client node as the search query, while the search query is being entered by the user. As the user further enters the search query, a new speculative search result may be provided to the user.

One embodiment selects from a set of query-suggestion pairs a first query and a subset of query-suggestion pairs that each has the first query as its query; computes a Log Likelihood Ratio (LLR) value for each query-suggestion pair from the subset of query-suggestion pairs; ranks the subset of query-suggestion pairs according to their respective LLR values; removes from the subset of query-suggestion pairs all query-suggestion pairs whose LLR values are below a predetermined LLR threshold; computes a Pointwise Mutual Information (PMI) value for each remaining query suggestion pair from the subset of query-suggestion pairs; removes from the subset of query-suggestion pairs all query-suggestion pairs whose PMI values are below a predetermine PMI threshold; and constructs a ranked set of suggestions for the first query, wherein the ranked set of suggestions comprises one or more suggestions of the remaining query-suggestion pairs from the subset of query-suggestion pairs.

Search assist powered by session analysis

Performing an ordered search of different databases in response to receiving a search query and without receiving any additional user input

Applying a bias when determining a suggested search query. Examples of biases that can be applied include, but are not limited to, temporal biases and monetization biases. Temporal biasing involves increasing a weight associated with a search query, based on a temporal attribute associated with the query. Search queries may also have associated with them a parameter such as frequency, count, etc. One of these parameters may serve as a weight. In order to determine suggested search queries, the temporal attribute is used to modify or bias the parameter (e.g., frequency parameter). Thus, the weight of the search query is modified to temporally bias the query. The suggested search queries are determined based on the biased parameter.

Biasing queries to determine suggested queries

Providing an alternative search query to a predicted search query is disclosed herein. A search query is received from a client node. Prior to receiving an indication from the client node that the search query is completely formed, the following steps are performed: 1) a predicted search query is determined by predicting what the search query will be when completed; and 2) an alternative search query that differs from the predicted search query is determined based on the predicted search query. The alternative search query is provided to the client node. The alternative search query may be something that the user search query is unlikely to complete to. For example, in response to the user entering a search query of 'brittany sp', an alternative search query with a spelling suggestion of 'britney spears' is determined and provided to the user.

Alternative search query prediction

Speculative search result based on a not-yet-submitted search query

A method and apparatus for providing a highly interactive user interface to a search engine using user input prediction is disclosed. As the user enters a portion of a query in a query field, the portion of the query is sent over a network to a query predictor. The query predictor determines a set of one or more predicted queries that correspond to the portion of the query. A particular predicted query is selected from the set of predicted queries based on selection criteria. The particular predicted query is processed by a search engine to obtain search results. The set of predicted queries and the search results are sent to the user and displayed on the user interface. The user may choose a different predicted query in the set of predicted queries or continue entering characters, both of which may result in different search results being displayed.

Interactive search engine

Growing community of inventors

Creston, Canada

Richard Kasperski