=Paper=
{{Paper
|id=Vol-1179/CLEF2013wn-INEX-MishraEt2013
|storemode=property
|title=SPAR-Key: Processing SPARQL-Fulltext Queries to Solve Jeopardy! Clues
|pdfUrl=https://ceur-ws.org/Vol-1179/CLEF2013wn-INEX-MishraEt2013.pdf
|volume=Vol-1179
|dblpUrl=https://dblp.org/rec/conf/clef/MishraGT13
}}
==SPAR-Key: Processing SPARQL-Fulltext Queries to Solve Jeopardy! Clues==
https://ceur-ws.org/Vol-1179/CLEF2013wn-INEX-MishraEt2013.pdf
SPAR-Key: Processing SPARQL-Fulltext Queries
to Solve Jeopardy! Clues
Arunav Mishra1 , Sairam Gurajada1 , and Martin Theobald2
1
Max Planck Institute for Informatics, Campus E1.4, Saarbrücken, Germany
2
University of Antwerp, Middelheimlaan 1, 2020 Antwerp, Belgium
Abstract. We describe our SPAR-Key query engine that implements indexing,
ranking, and query processing techniques to run a new kind of SPARQL-fulltext
queries that were provided in the context of the INEX 2013 Jeopardy Task.
1 Introduction
The LOD track of INEX 2012 introduced the new Wikipedia-LODv1.1 collection that
combined highly structured semantic data and unstructured (or semi-structured) textual
data with the goal of improving IR-tasks (adhoch search task and faceted search task)
and Question-Answering tasks (Jeopardy task). The entity-centric collection comprised
XML-ified documents, coined Wiki-XML documents, as the basic unit of data, where
each documents corresponded to a Wikipedia entity, and combined semantic (RDF)
data from DBpedia and YAGO2 Knowledge Bases, and textual data from the Wikipedia
article that describe the entity. The Jeopardy task, defined on the Wikipedia-LODv1.1
collection as a part of the LOD track, evaluated retrieval techniques over a unique query
benchmark of 90 queries, given in the new SPARQL-FT format. The SPARQL-FT
queries of the benchmark represented a translation of Jeopardy-style Natural Language
(NL) questions into a combination of W3C standard SPARQL and traditional keyword
queries [7]. In INEX 2013, the same efforts have been continued for the LOD track,
however, with a new and more complete version of the collection, Wikipedia-LODv2.0,
and a larger query benchmark for the Jeopardy task with 105 queries.
In this paper, we introduce a query engine, called SPAR-Key, which we developed
in the context of our participation in the Jeopardy Task of the INEX 2013 Linked Data
Track. This work has been a continuation of our attempts in the INEX 2012 Jeopardy
task [1]. At a high level, the SPAR-Key engine translates a SPARQL-FT query into a
conjunctive SQL query and processes it over a relational database schema. We delin-
eate three SPARQL-FT-to-SQL translators as a part of the SPAR-Key engine, namely
SPAR-Key Supremacy, SPAR-Key Ultimatum-No Phrases, and SPAR-Key Ultimatum-
Phrases, and submit runs based on each variant. In addition, we motivate and analyze the
translation strategies, and investigate the improvements in the quality of results against
the official evaluations released by the INEX community.
2 Data Storage in Relational-DBMS
Data management and storage becomes a critical in designing a query processor. Think-
ing of scalable data storage, the most common option is a Relational Data Management
Attribute Value
Column Type Number Of Rows 200187000
N3ID NUMBER Blocks 4288182
Subject VARCHAR2(1024) Distinct Rows Subject 18272256
Predicate VARCHAR2(1024) Distinct Rows Object 26873856
Object VARCHAR2(1024) Distinct Rows Predicate 18310
Table 1. Table schema the Table 2. Table Summary of the
DBpediaCore table DBpediaCore table
System (relational-DBMS). Though we also explored other approaches like file sys-
tems, graph-databases, etc., a relational-DBMS based approach seemed to be a viable
option due to the ease of implementation and re-use of the optimization techniques gen-
erally adopted by the database community. Designing a respective database schema as
our storage model suffices our need, as we aim to investigate the effectiveness and not
maximize the efficiency of the query processing.
2.1 Storage Model for Structured Data
The RDF data can be commonly perceived as a collection of triples of the form Subject
(S), Predicate (P) and Object (O) or, Entity, Property and Value (in entity-relationship
terminology). In the last decade, there have been many perspectives put forward by dif-
ferent research communities to manage RDF data. We identify the three most important
in order to find the best suited data storage model for our collection, namely a rela-
tional perspective, an entity perspective and a graph-based perspective [4], and finally,
we adopt the relational perspective for our system.
Storing RDF Data as a Single Relational Table. The vertical representation under
the relational perspective enables us to view RDF data as a large collection of triples
containing SPO components. By assuming such a serialization of RDF data into a flat
relational table, simply means that a given structured query in SPARQL has to be trans-
lated into a SQL query and issued to the relational-DBMS.
As our storage back-end, we use the Oracle 11g relational-DBMS to store the RDF
data that we imported from the data dumps of DBpedia and YAGO2. Table 1 shows the
schema of the table that stores the entire structured part of the collection. We call this
table DBpediaCore table and from here on, we refer to the table with this name.
Creation of the DBpediaCore Table. To parse the structured part of the collection
constituting RDF facts, we make use of current Linked Open Data dumps for DBpedia
(v3.8) and YAGO2, which are available from the following URLs:
– DBpedia v3.8:
http://downloads.dbpedia.org/3.8/en/
– YAGO2 core and full dumps:
http://www.mpi-inf.mpg.de/yago-naga/yago/
Column Type Number Of Rows 1688869800
Entity ID VARCHAR2(1024) Blocks 15228747
Term VARCHAR2(1024) Distinct Rows Entity ID 6180437
Score NUMBER Distinct Rows Term 1576558
Table 3. Table schema of the Keywords Table 4. Table Summary of the
table Keywords table
The N-Triple (.nt) format of the dumps are downloaded and are bulk-loaded into the
Jena RDF engine. We make use of the Apache Jena TDB to bulk load the triples into
the engine. Jena TDB build its own indexes over the data that can be used to efficiently
process SPARQL queries over an RDF store. However, we use the JAVA interface pro-
vided by the Jena TDB to traverse over all the triples and further bulk load them into
a relational table. Usage of Apache Jena TDB is not strictly necessary for building the
triple store in the relational-DBMS as any standard RDF parser could do the job. Our
use of Jena was solely to exploit its fast bulk loading mechanism. Table 2 shows some
statistics of the DBpediaCore table.
2.2 Storage Model for Unstructured Data
Traditionally, inverted indices are used as core data structures in keyword-based re-
trieval systems. Abstractly, these inverted indices facilitate efficient fulltext searches
and retrieval of most relevant or best matched documents to a given keyword query.
Essentially, inverted indices map every term (as keys) in the corpus to a set of docu-
ments with a similarity score that is generated by a scoring function. We use a similar
approach however by first generating and later translating such an inverted index into a
relational table.
Storing Textual Data in a Single Relational Table. Commonly keyword-based re-
trieval systems at least store a map or inverted index mapping every term to the docu-
ments in the corpus. This can be viewed as relational data and can be stored in relational-
DBMS. Realizing this, we create a relational table called Keywords table to store all
the term-entity pairs extracted from the Wikipedia fulltext collection. From here on we
refer to the table with this name. The schema of this table is shown in Table 3.
The Entity ID column essentially stores the Uniform Resource Identifier (URI)
of the DBpedia entities. Since in our entity-centric collection, every document corre-
sponds to a DBpedia entity, we prefer to use the prefixes defined by DBpedia to rep-
resent these entities, for example htttp://dbpedia.org/resource/entity.
Every tuple of the Keywords table represents a term mapped to DBpedia entity and a
similarity score to the entity’s Wikipedia page.
Creation of the Keywords Table. We employee a regular SAX parser to parse the
XML articles whose general XML structure is still based on that of the original articles.
That is, these articles contain a meta-data header with information about the ID, au-
thors, creation date and others, usually also an infobox with additional semi-structured
Index Name Attributes
DBpediaIDX Obj (Object,Subject,Predicate)
DBpediaIDX Sub (Subject,Predicate,Object)
DBpediaIDX Prd (Predicate,Object,Subject)
Table 5. Schema of the DBpediaCore table
information consisting of attribute-value pairs that describe the entity, and of course rich
text contents consisting of unstructured information and more XML markup about the
entity that is captured by such an article. Our keyword indexer uses the basic function-
ality of TopX 2.0 [5], which includes Porter stemming, stopword removal and BM25
ranking, but stores the resulting inverted lists for keywords into the Keywords rela-
tional table instead of TopX proprietary index structures. Table 4 shows some statistics
of the Keywords table.
2.3 Indexes on the Relational Tables
We note that solving a complex SPARQL query, with one or more logical joins of
triples, employs multiple self joins of DBpediaCore table and also Keywords table.
This is a performance killer due to the colossal size of the tables. Relational-DBMS
systems provide a standard optimization by facilitating index creations over a relational
table.
Many approaches motivate to build multiple indexes for different permutations and
combinations of the triple components. For example RDF-3x [6] creates 15 indexes
over a triple store. However, we observe that we already achieve decent efficiency with
three non-unique, visible and composite indexes on DBpediaCore table as shown
in Table 5. Similarly we create two non-unique, visible and composite indexes on the
Keywords table. The two indexes are built with the consideration that queries are
issued with conditions on the Entity ID column and Term column while the Score
column is used for purposes of ranking. Table 6 describes these two indexes built over
the Keywords table.
2.4 Keyword Ranking: Okapi BM25
In this section, we present the scoring model used to generate the per term-entity scores
stored in the third column of the Keywords table. There many well studied state-of-art
scoring functions that work well for a definite setting and it would be difficult to claim
a generic ranking function that is optimal. For our data, we select a variant of the Okapi
BM25 [2] scoring function with parameters k=1.2 and b=2.0, which works well for the
Index Name Attributes
Keywords Entity IDX (Entity ID, Term, Score)
Keywords Term IDX (Term, Entity ID, Score)
Table 6. Indexes built over the Keywords table
fulltext searches involved in our query processing. The exact BM25 variant we used for
ranking an entity e by a string of keywords S in an FTContains operator is given by
the following formula:
X (k1 + 1) tf (e, ti ) N − df (ti ) + 0.5
score(e, FTContains(e, S)) = · log
K + tf (e, ti ) df (ti ) + 0.5
ti ∈S
len(e)
with K = k1 (1 − b) + b
avg{len(e ) | e0 in collection}
0
where,
1) N is the number of XML articles in Wikipedia LOD collection.
2) tf (e, t) is the term frequency of term t in the Wikipedia LOD article associated
with entity e.
3) df (t) is the document frequency of term t in the Wikipedia LOD collection.
4) len(e) is the length (sum of tf values) of the Wikipedia LOD article associated
with entity e.
We used the values of k1 = 2.0 and b = 0.75 as the BM25-specific tuning parame-
ters (see also [2] for tuning BM25 on earlier INEX settings).
2.5 Entity Ranking
We realize that ranking entities in context of a given structured query becomes a difficult
challenge. As a simple approach, we consider it to be reasonable to carry over the
aggregated scores of the entities obtained from fulltext searches that are performed for
associated fulltext constraints on the entity. However, there may arise a special case
where in a SPARQL-FT query, entities do not have fulltext constraints. In such a case
all the entities that satisfy the semantic structure defined by the query triple (or triples)
pattern, become candidates to either become the final answer, or for further processing.
In such cases, we give a default score of 1 to these entities. A more elaborate discussion
is presented in [1].
3 SPAR-Key Supremacy
In this section, we introduce the first variant of our query engine, called SPAR-Key
Supremacy. In addition to the basic ranking methodology discussed in the last section,
this system implements a SPARQL-FT-to-SQL translator that: 1) uses SQL Joins over
simple “AND conditions” in conjunctive query processing, 2) materializes temporary
tables and sub queries to represent intermediate results to improve efficiency, 3) uses
a simple selectivity estimation technique to decide join order of temporary tables, and
4) includes additional query optimizations to force the optimizer to follow the decided
join order.
3.1 Materializing Temporary Tables
One big conjunctive query forces the Oracle optimizer to choose from a very large
number of possible query execution plans, and it often chooses an inefficient plan.
Thus, to prevent the optimizer from taking such inappropriate decisions, we materi-
alize temporary tables by separately joining the Keywords table instances and the
DBpediaCore table instances. This strategy acts as a strong hint for the optimizer.
The optimizer selects better query plans for the smaller intermediate queries and store
their results into temporary tables which are later joined together to retrieve the final
result.
3.2 Evaluating the Join Order and Forcing Orders via Optimizer Hints
There are some simple techniques by which we can determine the join order of the
tables. One such technique is to maintain an Inverse Document Frequency (IDF) index
containing the most frequent terms that occur in the collection. This index follows a
very simple layout of a key-value pair, where a key is a term and the value is it’s IDF.
A frequent term will have lower IDF and hence a select query on the Keywords table,
with the term as constraint, will return a larger result set. At the same time, if a term
is absent in the feature index, it can be assumed to be infrequent. Every instance of
the Keywords table can now be joined in increasing order of the IDF values of their
respective terms, thus ensuring the smaller tables to be joined first.
In our case, since we use the Oracle relational-DBMS as our back-end, it provides
a functionality by which the joining of intermediate results can be enforced on the
Oracle optimizer. This is achieved by adding optimizer hints to the queries. Of the made
available hints by Oracle to guide the query optimizer, we identify that the Ordered
hint could force the joining of tables in the determined order while preserving the logics
of the join condition in the original query. Thus our query translator automatically adds
this hint in the translated SQL queries.
3.3 SPAR-Key Supremacy: The Rewriting Algorithm
We can now develop an overall rewriting algorithm by putting together all the afore
mentioned steps.
1. Load the features index containing frequent terms and their IDF values into main
memory.
2. Tokenize and stem the FTContains fulltext conditions and decide the order of
joins among the keywords from the Features index.
3. Create temporary Keysi tables for each fulltext condition: these contain the results
of the OUTER joins over the Keywords table constrained by the terms.
4. Create temporary T abi tables for each triplet pattern. These contain the results of
the INNER joins over the DBpediaCore table which are additionally joined with
Keysi temporary tables for each FTContains fulltext condition in the query.
5. Assign a default score of 1 to all triples without any fulltext condition.
6. Formulate the main select query that combines the T abi temporary tables via an
INNER joins; the join logic is based on the joins given in the original SPARQL
query.
7. Finally, drop the temporary tables Keysi and T abi .
4 SPAR-Key Ultimatum
In this section we introduce the second and third variant of the query engine, called the
SPAR-Key Ultimatum No Phrases and the SPAR-Key Ultimatum Phrases. In fact,
these are implementation-wise the same with only one additional component called
Phrase Search (discussed in later in Section 4.4) activated in the latter. Thus we describe
them together in this section.
In addition to the basic ranking and efficiency improvement methodologies dis-
cussed so far, this system implements a SPARQL-FT-to-SQL translator that: 1) uses
Class Selection to prune out false positives, 2) exploits the structure of a give query
to identify additional constraints, 3) incorporates URI Search as a basic entity disam-
biguation tool, and 4) extracts noun phrases and performs proximity search to improve
ranking.
4.1 Class Selection
We note that additional contextual information can be used to prune out irrelevant enti-
ties before performing intermediate INNER joins or OUTER joins, to create temporary
Tab or Keys tables (described in the 3.3). For example, entity has a
type . It is not hard to see that context or type of an
entity can be derived from the DBpedia class to which it belongs. Due to neat hierarchy
of classes defined in DBpedia, we can safely obtain the classification of entities. Using
this knowledge we can logically partition the RDF data graph based on the classes so as
to reduce the search space in the intermediate steps. Figure 1 shows a snap-shot of the
class structure defined in DBpedia found at http://mappings.dbpedia.org.
We are only concerned with the DBpedia and YAGO2 classes identified by prefix:
and
.
The property definition in a class specifies a signature. To understand this, let us
consider an example shown in Figure 2. This example shows the properties defined
in the class with their signature. In the above example, one can see
a property defined with a Domain, and Range
which are DBpedia classes. Thus if this property should occur
as a Predicate in a triple pattern then all the entities classified as should
occur as the Subject of the triple and all the entities classified as
should occur as the Object of the triple. This forms an important observation to de-
rive the classes of the entities that can occur in a triple pattern of a given query. We
can therefore reduce the search space only to those entities that belong to the marked
classes.
Fig. 1. Snapshot of DBpedia class hierarchy Fig. 2. Snapshot of DBpedia class hierarchy
[http://mappings.dbpedia.org] [http://mappings.dbpedia.org]
To implement this, we create two indexes that store the Predicates along with their
Domains and Ranges separately. The first index is called the PredicateDomainIDX
and the second index is called the PredicateRangeIDX. Table 7 depicts the schema
of these indexes. These indexes facilitate the class markings of the Subjects and Objects
on-the-fly while processing a query. These indexes prove not to be very large in size
and can hence easily be loaded into the main memory.
4.2 Exploiting the Query Structure
Interpreting a structured query as a basic graph pattern, leads us to observe two common
kinds of query patterns. They are commonly known as a Chain pattern query and a Star
pattern query [6].
– Chain pattern is where the Object of the first triple pattern is the Subject of the next
triple pattern, again with given Predicates. Figure 3 shows a generic illustration of
a chain pattern.
– Star pattern is where multiple triple patterns with different Predicates share the
same Subject. These are used to select specific subjects. Figure 4 shows a generic
illustration of a star pattern.
The query pattern can be used to derive classes of entities even though in a triple
pattern a Predicate is not specified by a literal. To see this let us consider an exam-
ple query as shown in Figure 4. By analyzing the Predicate of the first triple pattern
( ?a.), i.e., , we can mark the Object (?a)
Index Name Attributes
PredicateDomainIDX (Predicate, Domain)
PredicateRangeIDX (Predicate, Range)
Table 7. Schema of the PredicateDomainIDX and PredicateRangeIDX indexes
Index Name Attributes
URI IDX (Entity ID, Term)
Table 8. Schema of the URI index
of to be class which is specified by the Range of the Predicate signature ; and
then by analyzing the second Predicate , we can mark the Subject (?c) of
the second triple pattern (?b ?a.) to be of class . Similar
analogies can be drawn for chain query patterns also.
4.3 Search on the URI Index
Document titles are considered as an important feature by IR systems among other
features like content, context, etc. These document titles tend to summarize the major
context of the articles. Following this notion, we also find that most of the entity de-
scriptions or key-concepts in a fulltext condition map to the surface forms of the entity.
In our collection these surface forms tend to occur as document titles. Following this
idea, we create an additional URI index other than the above described Keywords
indexes that mainly stores the surface forms of the articles. For a fulltext condition,
SPAR-Key Ultimatum performs an additional look-up on the URI index and then per-
forms an OUTER join with the results of a fulltext search on the article content. By
doing this, we include the entities that are missed by a fulltext search on the content.
Also the scores of the entities that are found by both the searches are boosted. Table 8
shows the schema of the URI index used by SPAR-Key Ultimatum.
4.4 Phrase Search
We observe that the in most of the Jeopardy-style NL questions, the occurrence of the
clues are in form of phrases, for example “King of Pop”, “Don’t be Evil”, etc. In ad-
dition, we also observe that other clues that mostly describe an entity (in the associated
fulltext conditions), contain nouns that most likely to occur in close proximity in the
textual data corresponding to the entities, for example, “Christian church founder”,
“video sharing”, etc. Thus we identify these noun phrases as additional features to
improve the ranking of the entities. The main idea is to perform a basic Natural Lan-
guage Processing on the Jeopardy-style questions to extract all the noun phrases and
Fig. 3. Chain query structure Fig. 4. Star query structure
then boost the scores of the entities that contain these noun phrases. To perform the
noun phrase extraction, the SPAR-Key Ultimatum uses the Stanford NLP Core [3] as a
black box and then automatically formulates a proximity keyword query by setting the
slope between the keywords as the average length of a sentence in Wikipedia. It then
performs an OUTER join of the results from the proximity search to the the results of
the fulltext search (by assuming independence between keywords), boosting the scores
of the entities that are returned by both.
4.5 SPAR-Key Ultimatum: The Rewriting Algorithm
1. Load the features index containing frequent terms and their IDF values into main
memory.
2. Tokenize and stem the FTContains fulltext conditions and decide the order of
joins among the keywords from the features index.
3. Analyze the Predicates in each triple and mark the bounded variables with their
class. The domain of the Predicate will be marked for the Subject and Range will
be marked for the Object. Thus we obtain a set:
classSet = {{variable1 , class1 }, {variable2 , class2 }, {variable3 , class3 }, ...}.
4. Exploit the query structure to find most selective class for an entity place-holder
and update the classSet.
5. Create Keys SearchU rii tables containing results form a search on the entity URI
for each term.
6. Create temporary Keysi tables for each fulltext condition: these contain the results
of the OUTER join over the Keywords table instances constrained by the terms.
Also add the class constraint on the bounded entity with the corresponding class
value in classSet.
7. Create temporary Keys Search F inali tables by performing a FULL OUTER
JOIN on Keys SearchU rii and Keysi .
8. Create temporary T abi tables for each triplet pattern. These contain the results of
the INNER join over the DBpediaCore table instances which are addition‘ally
joined with Keys Search F inali temporary tables for each FTContains full-
text condition in the query. Also add the class constrains to the variables by select-
ing the class values from the classSet.
9. Assign a default score of 1 to all triples in absence of a fulltext condition.
10. Formulate the main select query that combines the T abi temporary tables via an
INNER join; the join logic is based on the joins given in the original SPARQL
query.
11. Finally, drop the temporary tables Keys SearchU rii , Keysi , Keys Search F inali
and T abi .
5 Evaluations
In this section we provide experimental evaluation of our SPAR-Key query processor
over the Wikipedia-LOD collection. The evaluation studies the effectiveness of answer-
ing a Jeopardy-style Natural Language question translated into a SPARQL-FT query
with query processing techniques proposed in this paper.
5.1 Experimental Setup
Preprocessing of the data collection before storing into Oracle 11g relational-DBMS,
is done on a machine with Intel Xeon processor at 2.79 GHz. The machine has a main
memory of 64 GB and secondary memory of 1 TB. To generate runs for the benchmark
queries, we use a personal computer with Intel Core i3 processor at 3.30 GHz. This
machine has a main memory of 8 GB and secondary memory of 200 GB. This machine
is running a 64bit-Windows operating system.
5.2 Measures
We use standard TREC metrics to measure the performance of all the runs. To com-
pare our SPAR-Key engine variants we use the Mean-Average-Interpolated-Precision
(MAiP), Precision at K (P@K), specifically P@5, P@10, P@15 and show plotting of
the Average-interpolated-Precision values (AiP) at 11 standard points. We also per-
form a QA style evaluation with the Mean-Reciprocal-Rank (MRR) and Normalized-
Discounted-Cumulative-Gain (NDCG) specifically, NDCG@5, NDCG@10 and NDCG
@15. For further details on the metrics used to evaluate the run please refer to [8].
5.3 Experimental Runs
In this section we analyze our query engine based on the official evaluation results
presented by the INEX. Table 9 shows the official INEX results over the Jeopardy top-
ics. To generate the results, top-20 results from each run were pooled and assessed by
crowd sourcing through the Crowdflower platform 3 . For further details on the evalu-
ation procedure please refer to the Overview paper of the INEX 2013 LOD track. We
note that we were the only group that participated in the Jeopardy task this year and
hence we cannot compare our results to any other competitor. However, we compare
our own variants of the query processor based on the three runs (one from each variant)
submitted to the INEX.
We identify four classes of queries in the benchmark that: 1) target single entity as
the only correct answer, 2) target combination of entities as the only correct answer, 3)
target list of entities as correct answer and 4) target list of combination of entities as
correct answer. In this evaluation we present the combined results of our engine over
all the queries.
From the evaluation results we observe that by activating the phrase search compo-
nent (described in Section 4.4) we obtain the best results in terms of MRR and NDCG
values. This clearly supports our choice of recognizing the noun phrases as valuable fea-
tures for the query processing. In addition, we find that the simpler Supremacy variant
gives better performance than the Ultimatum-No phrase (though not significantly bet-
ter) as the URI Search (described in Section 4.3) which is activated in the Ultimatum-No
phrase becomes an overkill for most of the queries.
3
https://crowdflower.com/
MPIUltimatum Phrase MPIUltimatum NoPhrase MPISupremacy
MAiP 0.7491 0.701 0.719
MRR 0.7671 0.7358 0.7539
NDCG@5 0.7723 0.7307 0.7393
NDCG@10 0.7864 0.7347 0.7598
NDCG@15 0.7968 0.7484 0.7728
AiP@1% 0.7804 0.7411 0.7669
AiP@10% 0.7804 0.7411 0.7669
AiP@20% 0.7804 0.731 0.7653
AiP@30% 0.7804 0.731 0.763
AiP@40% 0.7772 0.7255 0.7468
AiP@50% 0.7737 0.7232 0.7417
AiP@60% 0.7337 0.6803 0.6991
AiP@70% 0.7245 0.6771 0.6952
AiP@80% 0.7223 0.6747 0.685
AiP@90% 0.7208 0.673 0.6817
AiP@100% 0.7208 0.6662 0.6694
Table 9. Jeopardy Task retrieval results.
6 Conclusion
We presented an approach for storing structured RDF data and unstructured data in
relational database. We also presented the necessary indices required to efficiently pro-
cess queries over the relational schema. Our approach converts a SPARQL query with
fulltext conditions into unions of conjunctive SQL queries by materializing temporary
tables. These temporary tables store intermediate results from inner or outer joins over
our relations, based on given conditions in the query. We also presented a simple yet ef-
fective way to rank entities by translating scores from keywords. In addition, we showed
three variants of the query processor, each following different query processing strate-
gies by recognizing a different set of features and score boosting to obtain the entity
ranking. Finally we compared the variants based on the official evaluation released by
the INEX and underline some of the key advantages and disadvantages of each query
processing strategies. As a future work we would like to focus on the efficiency of the
query processing and shifting the storage model from relational-DBMS to file system.
Also we are keen in designing an automatic translator for translating a Jeopardy-style
NL question into a SPARQL-FT query which can be then processed by our current
system.
References
1. Arunav Mishra and Sairam Gurajada and Martin Theobald. Running SPARQL-Fulltext
Queries Inside a Relational DBMS. In CLEF (Online Working Notes/Labs/Workshop), 2012.
2. Clarke, Charles L. A. Controlling overlap in content-oriented XML retrieval. In SIGIR, 2005.
3. M.-C. De Marneffe and C. D. Manning. Stanford typed dependencies manual. URL http://nlp.
stanford. edu/software/dependencies manual. pdf, 2008.
4. Luo, Yongming and Picalausa, François and Fletcher, George H. L. and Hidders, Jan and
Vansummeren, Stijn. Storing and Indexing Massive RDF Datasets Semantic Search over the
Web. In Semantic Search over the Web, Data-Centric Systems and Applications. 2012.
5. Martin Theobald and Ablimit Aji and Ralf Schenkel. TopX 2.0 at the INEX 2009 Ad-Hoc
and Efficiency Tracks. In INEX, 2009.
6. Neumann, Thomas and Weikum, Gerhard. RDF-3X: a RISC-style engine for RDF. Proc.
VLDB Endow., 1, 2008.
7. Qiuyue Wang and Jaap Kamps and Georgina Ramirez Camps and Maarten Marx and Anne
Schuth and Martin Theobald and Sairam Gurajada and Arunav Mishra. Overview of the INEX
2012 Linked Data Track. In CLEF (Online Working Notes/Labs/Workshop), 2012.
8. Qiuyue Wang and Jaap Kamps and Georgina Ramirez Camps and Maarten Marx and Anne
Schuth and Martin Theobald and Sairam Gurajada and Arunav Mishra. Overview of the INEX
2013 Linked Data Track. In CLEF (Online Working Notes/Labs/Workshop), 2013.