Similarity queries are widely used in applications where users need to find records that satisfy a similarity predicate, while exact matching is not sufficient. These queries are especially important for social and Web applications, where errors, abbreviations, and inconsistencies are common. As an example, we may want to find all the movies starring Schwarzenegger, while we don't know the exact spelling of his last name (despite his popularity in both the movie industry and politics :-)). As another example, we want to find all the Facebook users who have similar friends. To meet this type of needs, AsterixDB supports similarity queries using efficient indexes and algorithms.
AsterixDB supports [http://en.wikipedia.org/wiki/Levenshtein_distance edit distance] (on strings) and [http://en.wikipedia.org/wiki/Jaccard_index Jaccard] (on sets). For instance, in our [https://code.google.com/p/asterixdb/wiki/AdmAql101#ADM:_Modeling_Semistructed_Data_in_AsterixDB TinySocial] example, the friend-ids of a Facebook user forms a set of friends, and we can define a similarity between two sets of friends. We can also convert a string to a set of grams of a length q (called "n-grams") and define the Jaccard similarity between the two gram sets of the two strings. Formally, the "n-grams" of a string are its substrings of length "n". For instance, the 3-grams of the string schwarzenegger are sch, chw, hwa, ..., ger.
AsterixDB provides [https://code.google.com/p/asterixdb/wiki/AsterixDataTypesAndFunctions#Tokenizing_Functions tokenization functions] to convert strings to sets, and the [https://code.google.com/p/asterixdb/wiki/AsterixDataTypesAndFunctions#Similarity_Functions similarity functions].
The following [https://code.google.com/p/asterixdb/wiki/AsterixDataTypesAndFunctions#edit-distance query] asks for all the Facebook users whose name is similar to Suzanna Tilson, i.e., their edit distance is at most 2.
use dataverse TinySocial;
for $user in dataset('FacebookUsers')
let $ed := edit-distance($user.name, "Suzanna Tilson")
where $ed <= 2
return $user
The following [https://code.google.com/p/asterixdb/wiki/AsterixDataTypesAndFunctions#similarity-jaccard query] asks for all the Facebook users whose set of friend ids is similar to [1,5,9], i.e., their Jaccard similarity is at least 0.6.
use dataverse TinySocial;
for $user in dataset('FacebookUsers')
let $sim := similarity-jaccard($user.friend-ids, [1,5,9])
where $sim >= 0.6f
return $user
AsterixDB allows a user to use a similarity operator ~= to express a similarity condition by defining the similarity function and threshold using "set" statements earlier. For instance, the above query can be equivalently written as:
use dataverse TinySocial;
set simfunction "jaccard";
set simthreshold "0.6f";
for $user in dataset('FacebookUsers')
where $user.friend-ids ~= [1,5,9]
return $user
In this query, we first declare Jaccard as the similarity function using simfunction and specify the threshold 0.6f using simthreshold.
AsterixDB supports fuzzy joins between two data sets. The following [https://code.google.com/p/asterixdb/wiki/AdmAql101#Query_5_-_Fuzzy_Join query] finds, for each Facebook user, all Twitter users with names "similar" to their name based on the edit distance.
use dataverse TinySocial;
set simfunction "edit-distance";
set simthreshold "3";
for $fbu in dataset FacebookUsers
return {
"id": $fbu.id,
"name": $fbu.name,
"similar-users": for $t in dataset TweetMessages
let $tu := $t.user
where $tu.name ~= $fbu.name
return {
"twitter-screenname": $tu.screen-name,
"twitter-name": $tu.name
}
};
AsterixDB uses a gram-based inverted index (called "n-gram") and efficient algorithms to support similarity queries. For a set of strings, we generate n-grams for each string, and build an inverted list for each n-gram that includes the ids of the strings with this gram. A similarity query can be answered efficiently by accessing the inverted lists of the grams in the query and counting the number of occurrences of the string ids on these inverted lists. The similar idea can be used to answer queries with Jaccard similarity. A detailed description of these techniques is available at this http://www.ics.uci.edu/~chenli/pub/icde2009-memreducer.pdf.
For instance, the following DDL statement creates such an index on the FacebookUser.name attribute using an inverted index of 3-grams. After the index is created, similarity queries with an edit distance condition on this attribute can be answered efficiently.
use dataverse TinySocial;
create index fbUserFuzzyIdx on FacebookUsers(name) type n-gram(3);
The number "3" in "n-gram(3)" is the length "n" in the grams. This index can be used to optimize similarity queries on this attribute using edit distance, or Jaccard queries on this attribute where the similarity is defined on sets of 3-grams. This index can also be used to optimize queries with the "contains()" predicate (i.e., substring matching) since it can be also be solved by counting on the inverted list of the grams in the query string.
AsterixDB also has an improved version of n-gram index called "Partitioned N-Gram Index". Its main idea is to partition the data into groups, and build an n-gram index for the records in each group. This partitioned index can be used to further improve query performance. The following is an example to declare such an index.
use dataverse TinySocial;
create index fbUserFuzzyIdx on FacebookUsers(name) type partitioned n-gram(3);