Geographical Queries Reformulation using Parallel FP-Growth for

Spatial Taxonomies Building

Omar El Midaoui

1,2

, Btihal El Ghali

and Abderrahim El Qadi

LRIT Associated Unit to the CNRST - URAC n°29, Faculty of Sciences, Mohammed V University in Rabat, Morocco

SmartiLab, Ecole Marocaine des Sciences de l'Ingénieur (EMSI), Rabat, Morocco

TIM, High School of Technology, Mohammed V University in Rabat, Morocco

Keywords: Information Retrieval, Parallel FP-Growth Algorithm, Machine Learning, Geographical Query

Reformulation, Spatial Entity, Spark, Big Data.

Abstract: Due to its specificities and hierarchical structure, a geographical query needs a special process of

reformulation by Information Retrieval Systems (IRS). This fact is ignored by most of web search engines.

In this paper, we propose an automatic approach for building a spatial taxonomy that models’ the notion of

adjacency that can be uses in the reformulation of the spatial part of a geographical query. This approach

exploits the documents that are in the top of the list of retrieved results when submitting a spatial entity, which

is composed of a spatial relation and a noun of a city. Then, a transactional database is constructed, considering

each document extracted as a transaction that contains the nouns of the cities sharing the country of the

submitted query’s city. The algorithm FP-Growth is applied to this database in his parallel version (PFP) in

order to generate association rules, that will form the country’s taxonomy in a Big Data context. Experiments

has been conducted on Spark and their results show that query reformulation based on the taxonomy

constructed using our proposed approach improves the precision and the effectiveness of the IRS.

1 INTRODUCTION

Most human activities are well located in the

geographical area. Thus, it is not surprising that a big

amount of web documents contain geographical

references. A study that was done on the Excite

search engine shows that between every five queries

there is one query which have a geographical context

(Sanderson and Kohler, 2004). Web users searching

for information that are spatially located often require

information, that are geographically specific, such

geographic terms in Web pages and user queries or

even user location (Jiang et al., 2018). However,

retrieval systems currently have limited support to

operationalize a user’s geospatial queries.

Geographic information deals with physical objects

that are in some cases hard to express with words and

that contain most of the time ambiguous terms. These

arguments prove the fact that it will be very useful for

search engines to take into account the spatial scope

of geographical queries.

The current search engines generally handle

queries by adopting a keyword matching approach

without inferring the geographical scope of the spatial

terms. Thus, when the name of a place is typed into a

typical search engine associated with a spatial

preposition (e.g. “near”), web pages that include that

name in the text will be retrieved but most likely, not

places that are close to that specified place.

In order to do a spatial analysis of text, the first

step is the annotation of spatial named entities.

Several techniques have proved their ability for

carrying out this annotation, such as the works of

(Rocío and Erick, 2010) and (Loustau, 2008) that has

elaborated it using external resources named

“gazetteers”. A gazetteer is a dictionary or geographic

directory whose inputs are names of places. Each

entry in the dictionary may be associated with

information such as belonging to one or more

administrative structures (town, region, country, etc.),

the physical characteristic (mountain, river, road,

etc.), statistical data, and a geometric representation

expressed in a geographic referential.

Other works proposes the categorization of these

spatial named entities after identification. Such as,

Buscaldi and Rosso that proposed a technique for

spatial named entities categorization using the

thesaurus Geo-WordNet (Buscaldi and Rosso, 2008).

El Midaoui, O., El Ghali, B. and El Qadi, A.

Geographical Queries Reformulation using Parallel FP-Growth for Spatial Taxonomies Building.

DOI: 10.5220/0009446603750381

In Proceedings of the 5th International Conference on Internet of Things, Big Data and Security (IoTBDS 2020), pages 375-381

ISBN: 978-989-758-426-8

375

Or Bouamor that exploited a document structure in

(Bouamor, 2009), using as a corpus the collaborative

encyclopaedia “Wikipedia”. In his work, the

identification of named entities is done using the title

and their categorization is based on the analysis of the

first sentence of the description part or the category

part at end of the article.

Different external resources have been also

created and used in the most recent works. Such as

the Geographical Information Retrieval model,

proposed by (Fang and Zhang, 2018), that simplifies

the process of user information acquisitions by

analyzing and extracting the attribute features, spatial

features and structure features of Geographic Markup

Language (GML) data. As defined by this work,

GML data is generated by artificial of special services

and stored semi-structured data using the Geographic

Mark-up Language (GML), which is a geographic

information coding specification that was

implemented based on the Extensible Mark-up

Language (XML) and standardized (ISO 19136-

2007).

In the other hand, some approaches aim more

particularly to the disambiguation of recognized

places names (Buscaldi, 2009). The ambiguity can be

understood as a word or a phrase that has many

meanings (Vargas et al., 2012). In this case, two types

of ambiguities are to treat (Einat et al., 2004): a

geo/non-geo ambiguity is when the entity has a non-

geographical meaning like the term “Turkey”, and a

geo/geo ambiguity that occurs when the named entity

refers to two different places as Rabat in Malta and

Rabat in Morocco.

A hybrid approach is proposed in (Gaio et al.,

2012) which, first landmark names of places but also

searches for these terms in ontological resources to

identify related terms, potentially geographic.

Domain-Specific taxonomies (Yangqiu et al., 2015)

are also playing an important role in many

applications for improving search results (Xueqing et

al., 2012) or helping with query reformulation as in

(Sadikov et al., 2010) and (Aloteibi, Mark Sanderson,

2014).

In this paper, we propose a geographical

taxonomy builder using the parallel FP-Growth

Algorithm (PFP) which inputs are text documents and

we complete the process by suggesting a query

reformulation approach for geographical queries. Our

contributions are also made in the step of

geographical and thematic entities separation giving

a geographical query, in order to reformulate the two

entities in a different manner.

This approach is tested using a collection that has

been created during our experimentations. This

collection contains 50 queries and 2500 documents.

We used 1500 documents, considering 30 retrieved

documents per city for the taxonomy building step, as

we used a list of the 50 most popular cities. In addition

of, 20 retrieved documents per submitted query in the

reformulation step evaluation (10 before and 10 after

reformulation). Thus, 2500 documents have been

used in total. The collection’s documents were

retrieved automatically using the google web services

whenever there was a need.

This article is organized as follows:

The section 2 is introducing our proposed

approach for the construction of a geographical

taxonomy of adjacency using the PFP algorithm,

while we explain our query reformulation technique

in section 3. The results of our experimentations are

presented in section 4. Finally, section 5 draws

conclusions and future works.

2 THE GEOGRAPHICAL

TAXONOMY BUILDER

A taxonomy consists of a number of names arranged

in a hierarchical system that describe a specific

domain (Enghoff, 2009) by a hierarchical structure. A

taxonomy starts from a general concept of a domain,

and associate to it the terms that describe this specific

domain more precisely while moving down in the

hierarchy.

In this work, we introduce an automatic approach

that builds a geographical taxonomy of adjacency. In

this aim, we exploit the best-ranked documents

retrieved using the search engine when submitting a

spatial part of a query, that contain the spatial relation

of adjacency and a noun of a city for which we are

constructing the taxonomy.

The proposed approach is based on the Parallel

FP-Growth algorithm. The geographical query model

used in this work considers two type of entities: the

Absolute and the Relative Spatial Entities (ASE and

RSE). The geographical named entities such as the

city of “London” are well-known named entities and

are defined as an ASE (Absolute Spatial Entity).

While complex spatial entities as “near London” are

labelled as an RSE (Relative Spatial Entities).

2.1 The FP-Growth Algorithm

The FP-growth technique (Han et al., 2004) is an

association rules Machine Learning algorithm, where

“FP” is the acronym of Frequent Pattern. Given as

IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security

376

input a dataset of transactions, the first step of this

algorithm is to compute item frequencies and identify

the most frequent items. Different from Apriori

algorithm (Najadat et al., 2013) (Al-Maolegi and

Arkok, 2014) designed for the same aim, by its

second step that uses a suffix tree structure, called FP-

tree, to encode transactions without the explicit

generation of candidate sets, which are usually

expensive to generate. After this step, the frequent

item sets are extracted from the FP-trees.

The FP-growth is a two phases algorithm. The

first phase consists on the construction of FP-Trees

and the second mines frequent patterns from the

generated FP-Trees.

The construction of an FP-Tree requires two scans

on the used database. The first scan permits the

selection of the frequent items that are then sorted

based on their frequency in descending order to form

a new structure caller F-list. The second scan

constructs the FP-Tree. First, the non-frequent items

are removed while reordering the database tuples

according to F-list. Then the reordered transactions

are inserted into the FP-Tree. The Input of the Growth

part of the algorithm is the constructed FP-Tree and

the value of minimum support threshold.

FP-Growth traverses’ nodes in the FP-Tree

beginning from the least frequent item in F-list. While

traversing each node, FP-Growth collects items on

the path from the node to the root of the tree. Those

collected items constitute the elements of the

conditional pattern base of the current item in F-list.

The conditional pattern base of an item is defined as

a small database of patterns that co-occur with this

item. Then FP-Growth creates small FP-Trees based

on the conditional pattern bases and re-executes the

algorithm recursively on the new FP-Trees until no

conditional pattern base can be generated.

2.2 The Parallel FP-Growth

The parallelized FP-growth work on distributed

machines (Lingling and Yuansheng, 2015). Its

partitions computation is done in such a way that each

machine executes an independent group of mining

tasks. This method of partitioning eliminates

computational dependencies between machines, and

thereby communication between them.

Given a transaction database DB, the PFP

algorithm’s steps are as follows:

 Sharding: splitting DB into successive parts and

storing those parts on n different machines. Each

resultant part is called a shard.

 Parallel Counting: counting the support values

of all items appearing in each shard. This step

permits to discover the items' vocabulary

implicitly, which is normally unknown for a

huge Database. The result of this step is an F-

list.

 Grouping Items: Considering I the set of

vocabulary discovered, splitting the |I| items

appearing in F-List into Q groups. The groups

list is called G-list, where each group is given a

unique group-id (gid). As F-list and G-list are

both small, this step can be executed on a single

node of the cluster in few seconds.

 Parallelizing: Selecting group-dependent

transactions on which the FP-Growth algorithm

is applied in order to build local FP-trees in

parallel and growth their conditional FP-trees

recursively.

 Aggregating: Aggregating the results generated

in Step 4 as our final result.

PFP distributes the growing FP-trees work based on

the transactions’ group. thus, this approach is more

scalable than a single-node implementation.

PFP is implemented in MLlib on Spark and it

takes three parameters: the minimum support

threshold to identify frequent item sets, the minimum

confidence for generating Association Rules and the

number of shards used to distribute the job.

2.3 Geographical Taxonomy of

Adjacency

Considering a database whose transactions are

documents and items are the cities of the country that

contain the city of the user query. We propose to build

a spatial taxonomy (Fig. 1) of adjacency based on the

PFP algorithm.

The documents that form the input transactional

database are restricted to the Absolute Spatial entities

contained in the documents. Thus, the items

considered are the ASEs.

After the application of the PFP algorithm,

starting from the capital of the country for which we

will built the taxonomy, the fusion of all the generated

FP-trees is forming our geographical taxonomy.

Figure 1: A two-level taxonomy for the ASE0.

Geographical Queries Reformulation using Parallel FP-Growth for Spatial Taxonomies Building

377

Validation Step. In this contribution, we propose also

a step of validation of each arc of the taxonomy. To

validate each arc in the aggregating step, we verify if

its two parts (the two ASEs that form this arc)

mutually generate each other in the FP-trees. For

example, ASE

involves ASE

and ASE

involves

ASE

then the arc is kept and this taxonomy evolves

to a two-level taxonomy as shown in figure 1.

Otherwise, ASE

has involved ASE

, but ASE

did

not involve ASE

so this arc has not been validated.

Thus, it has been removed from the taxonomy.

3 REFORMULATION OF A

GEOGRAPHIC QUERY USING

A TAXONOMY

In order to reformulate a geographical query, we first

separate the different components of the query based

on the approach of geographic information extraction

(GIE) proposed in (Sallaberry et al., 2007). This

approach utilizes a methodology of semantic

annotation for the detection of geographical markers:

first, the Absolute Spatial entity is detected and

annotated. Then the spatial entity (SE) is constructed

considering this ASE and a lexicon of spatial

relations. The remaining words of the query form the

thematic entity (TE).

In this step also, we proposed a contribution. We

made some modifications in the GIE approach cited

above based on a hypothesis.

Hypothesis. If the Spatial Relation is not present

in the query, the occurrence of an ASE does not mean

that the query has a geographical intent. For example,

a query containing “George Washington”. We can

also consider the example of the query searching for

“Hôtel de Paris”. In this context, the noun “Paris” is

the name of a hotel whose location is in Tangier,

Monte Carlo or Monaco.

After the separation of the different entities of the

user query, we continue applying the proposed

approach by interpreting the spatial relationship

contained in the spatial entity of the query. The

interpretation is done using a lexicon of adjacency

spatial relations. The process of reformulation

depends on the result of this interpretation.

If the spatial relation detected in the query is an

SR of adjacency, we reformulate the spatial part of

the query using the country’s taxonomy (Xueqing et

al., 2012).

Logically, a query that contains a relation of

adjacency means that the intent of the user is to

retrieve places that are around the ASE of his query.

Thus, we propose to eliminate the entire spatial entity,

and to replace by the direct child-nodes items (CNIs)

of the query’s ASE in the geographical taxonomy as

follows:

 User New Query = TE SR ASE

 Reformulated Query = expanded TE + “CNI 1”

or “CNI 2” or …

In the query resulted from the reformulation,

quotes are used to search for the desired place and not

separately search for the words that the place’s name

contains if the ASE is composed many terms (e.g. the

submission of New York unquoted, can lead the

search engine to search for New and York as two

independent terms). Moreover, the boolean operator

‘or’ is used, to ensure that the retrieval returns

documents that include for example "CNI 1" or "CNI

2" or both of them and so on for all the child-nodes

used to reformulate the query.

4 EXPERIMENTATION

RESULTS

To apply the proposed approach, we have used a

lexicon of spatial relationships, and a database of

validated ASEs associated with their countries.

In order to test and verify the performance of the

technique of taxonomy building proposed in this

work, we take our country Morocco as an example.

Thus, to be able to use the web pages created by

Moroccans themselves we perform our tests in

French. “Rabat”, the capital of Morocco is the ASE

that we took as a root of our taxonomy. The search

engine used in our experimentations is Google web

service.

We apply our method using transaction database

that is constructed by iterating on Morocco’s ASEs

list (a list of 50 cities and villages of Morocco). For

every ASE, we selected the thirty first web pages

retrieved when submitting a RSE containing the

current ASE.

As a pre-treatment step, we deleted accents from

the extracted documents to minimize the matching

gab between ASEs, due to different manners write

cities names by the persons who wrote the documents

contents. Because, the miss-matching problem arise

particularly in the case of nouns, which contain

accents.

Then, we varied the SR of the spatial entities

submitted to verify if the variation of the SR

influences the performance of the proposed approach.

The spatial relations used in this test step are as

follows:

IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security

378

Table 1: Spatial Relations.

Annotation Ex

ression

SR 1 à côté de

SR 2 à la

éri

hérie de

SR 3 à proximité de

SR 4 aux alentours de

SR 5 aux environs de

SR 6 les environs de

SR 7

rès de

First, the five top-ranked documents were

extracted for the ASE Rabat associated with every

spatial relationship of Table 1. A database (DB)

containing 35 transactions is constructed based on

these documents. The parallel FP-growth algorithm is

applied to this DB and then the association rules are

generated between Rabat and every Moroccan ASE

that co-occur with it in the DB.Ater that, we varied

the minsup from 0,2 to 0.8 (as shown in Table 2)

without the validation step for the rules extracted

using 2-frequent item sets. Later we computed the

error rate and the number of rules generated in every

case.

Table 2: The error rate and the number of rules generated

while varying the minimum support threshold and the

spatial relationship used for item sets containing the ASE

“Rabat”.

Error rate Number of

generated rules

RS\

Threshol

0,2 0,4 0,6 0,8 0,2 0,4 0,6 0,8

RS 1

72,73 28,57 33,33 0 22 7 3 1

RS 2

42 0 0 0 5 2 1 1

RS 3

25 0 0 - 4 1 1 0

RS 4

40 33,33 0 0 10 3 1 1

RS 5

33,33 0 0 0 9 2 2 1

RS 6

40 50 0 0 10 4 2 2

RS 7

0 0 0 - 6 6 1 0

From Table 2, we notice that using the

minsup=0,8 the algorithm does not return any results

in some cases otherwise it gives 1 or 2 answers. The

same for minsup=0,6 that do not exceed 2 correct

answers.

Regarding the value 0,2 it generally gives a high

error rate and sometimes returns a very high number

of responses up to 22 resulting ASEs in the case of

RS 1 with 6 correct adjacent ASEs only. Thus, we

favored the value of minimum support equal to 0,4

because it is the one that gives the best ratio between

a minimal error and an acceptable number of answers.

The next step of experimentations is done in order

to compare the cases where we use or not the

validation step for aggregating the generated FP-trees

in order to built the taxonomy of adjacency, based on

a minimum support of 0,4.

Table 3: The error rate and the number of correct rules

generated using the step of validation or not and using the

average of support between the two cases, varying the

spatial relation used for item sets containing the ASE

“Rabat” with a minsup of 0,4.

Error rate Number of correct

rules

Spatial

relation

WV UV AS WV UV AS

SR 1

28,57 0 0 7 2 2

SR 2

0 0 0 2 1 1

SR 3

0 - 0 1 0 1

SR 4

33,33 50 33,33 3 2 3

SR 5

0 0 0 2 2 2

SR 6

50 0 0 4 2 2

SR 7

0 0 0 6 5 6

WV: without validation, UV: Using validation, AS: Average support

Comparing the results using validation with the

results without validation, we note that the error rate

decreases when using the validation step, with the

exception of the SR 4 for which from 3 results

including 2 correct ASEs, validation has eliminated

one of the correct ASEs and kept the erroneous one.

Concerning the SR 3 we notice that the only ASE that

was resulted without validation was eliminated with

the step of validation. In general, we conclude that the

validation step reduces errors sufficiently.

To minimize the error rate while keeping as much

as possible of correct results (eliminate only the

erroneous ASEs by the validation step). We propose

to compute the average of the two supports of the

opposite rules (e.g. ASE1 → ASE2 and ASE2 →

ASE1). Table 4 shows that the result given by the case

of the average support solves the problems mentioned

above for the SR 3 and SR 4.

Comparing the seven spatial relations, we

promote the SR 7 “près de” which gives the most

interesting result with 0% error and six correct ASEs

as child nodes of Rabat’s taxonomy of adjacency.

Figure 2: A one-level taxonomy for Rabat using the spatial

relation “Près de”.

Salé

Temara

Kénitra

Bouznika

Skhirat

Harhoura

Rabat

Geographical Queries Reformulation using Parallel FP-Growth for Spatial Taxonomies Building

379

Using the favorable conditions represented above

we continue the construction of Morocco’s taxonomy

(as shown in figure 3) with 0,4 as a minimum support,

and using the average of support for validating links.

Figure 3: A two-level geographical taxonomy of adjacency

for Morocco.

For building this taxonomy, we had used 50

Moroccans ASEs and we were searching for the 30

first retrieved documents while submitting every

ASEs with the selected relationship. Thus, 1500

documents have been used in this test with a

minimum support of 0,4 and a minimum confident of

0,6. This test have been conducted in 0,85 seconds

due to the use of Spark, using a cluster of two nodes.

In order to evaluate the precision of the results of

our approach and confirm the results of the precedent

tests, we proposed 50 geographical queries that has

been submitted to the Google search services with

and without reformulation using the taxonomy

realized based on the PFP algorithm, and we

compared the values of the P@10 and the Mean

Average Precision of the two cases.

Table 4: The performance rate of the presented technique

according to the original queries.

Baseline P

10 MAP

Ori

inal

ueries 10,56% 12,92%

From Table 4, we notice that the approach

presented in this manuscript gives an interesting

improvement in the precision of the geographical

queries used in our experiments.

5 CONCLUSION AND FUTURE

WORKS

In this paper, we proposed a new method of

construction of geographical taxonomies of

adjacency using the parallel FP-Growth, and a

technique for reformulating geographical queries that

contain a spatial entity of adjacency. We have

conducted tests on the taxonomy builder method by

forming Morocco’s taxonomy of adjacency. During

our experimentations, we varied the minimum

support threshold and the used spatial relationship in

order to search for the parameters of the approach that

extract the most appropriate frequent item sets. Then

we constructed the Moroccan taxonomy using a

minimum support of 0.4 and a minimum confidence

of 0.6 and the SR number 7, because these conditions

gave the best results during our experiments. The

proposed technique of reformulation has been tested

on 50 queries, which have a geographical intent and

their thematic entities are from different fields. These

queries had been reformulated based on the spatial

taxonomy of adjacency. Finally, we compared the

results retrieved by the search engine before and after

the application of our technique using the evaluation

measures MAP and P@10. The results show that the

reformulation based on our proposed approach and

using a small number of reformulation terms has

improved the value of MAP significantly.

Considering the experimental results, we conclude

that the presented method is an efficient work that

permit to interpret and improve the results of queries

containing a spatial entity of adjacency.

As future work, we intend to propose a new

method of geographical query reformulation, based

on Big Data technologies and an in-depth analysis of

user’s behaviours through a study of a search engine’s

trace.

REFERENCES

Al-Maolegi, M., Arkok, B., 2014. An improved Apriori

algorithm for association rules, International Journal

on Natural Language Computing, Vol. 3, Issue 1, pp.

21-29, February 21-29.

Aloteibi, S., Sanderson, M., 2014. Analyzing geographic

query reformulation: An exploratory study, Journal of

the Association for Information Science and

Technology, Vol. 65, Issue 1 January, pp. 13-24.

Bouamor, H., 2009. Extraction des connaissances à partir

du Web pour la recherche des images géoréférencées,

in Proceedings of Conférence en Recherche

d’Information et Applications (CORIA), pp. 519-526.

Buscaldi, D., 2009. Toponym ambiguity in geographical

information retrieval, in Proceedings of the 32nd

international ACM SIGIR conference on Research and

development in information retrieval, SIGIR ’09,

ACM, New York USA, pp. 847-847.

Buscaldi, D., Rosso, P., 2008. Using GeoWordNet for

Geographical Information Retrieval, In proceedings of

CLEF, pp. 863-866.

Einat, A., Har'el, N., Sivan, R., Soffer, A., 2004. Web-a-

where: Geotagging Web Content, Proceedings of the

27th annual international ACM SIGIR conference on

Salé

Temara

Kénitra

Bouznika

Skhirat

Harhoura

Mehdia

Rabat

IoTBDS 2020 - 5th International Conference on Internet of Things, Big Data and Security

380

Research and development in information retrieval,

Sheffield United Kingdom, pp. 273-280.

Enghoff, H., 2009. What is taxonomy? – An overview with

myriapodological examples, Soil organisms, Vol. 81,

Issue 3, pp. 441–451.

Fang, C., Zhang, S., 2018. Geographic Information

Retrieval Method for Geography Mark-Up Language

Data, ISPRS International Journal of Geo-Information

Vol 7, Issue 3.

Gaio, M., Nguyen, V.T., Sallaberry, C., 2012. Typage de

noms toponymiques à des fins d’indexation

géographique, Revue Traitement Automatique des

Langues, Vol. 53, Issue 2, pp. 143-176.

Han, J., Pei, J., Yin, Y., 2004. Mining Frequent Patterns

without Candidate Generation: A Frequent-Pattern

Tree Approach, Data Mining and Knowledge

Discovery, January 2004, Vol. 8, Issue 1, pp 53–87.

Jiang, D., Cai, F., Chen, H., 2018. Location-Sensitive

Personalized Query Auto-Completion, in Proceeding of

the 10th International Conference on Intelligent

Human-Machine Systems and Cybernetics (IHMSC),

Zhejiang, China, 25-26 August, pp. 15-19.

Lingling, D., Yuansheng, L., 2015. Improvement and

Research of FP-Growth Algorithm Based on

Distributed Spark, In Proceeding of International

Conference on Cloud Computing and Big Data

(CCBD’2015), Shanghai, China, 4-6 November, pp.

105-108.

Loustau, P., 2008. Interprétation automatique d’itinéraires

dans des récits de voyage, PhD thesis, Université de

Pau et des Pays de l’Adour.

Najadat, H. M., Al-Maolegi, M., Arkok, B., 2013. An

improved Apriori algorithm for association rules,

International Research Journal of Computer Science

and Application, vol. 1, Issue 1, pp. 01-08.

Rocío, A.-M., Erick, L.-O., 2010. Geo information

extraction and processing from travel narratives,

Transforming the Nature of Communication, in

Proceedings of 14th International Conference on

Electronic, Helsinki Finland, pp. 363-373.

Sadikov, E., Madhavan, J., Wang, L., Halevy, A.Y., 2010.

Clustering query refinements by user intent, In

Proceedings of Proceedings of the 19th international

conference on World wide web, WWW’10, pp. 841–

850.

Sallaberry, C., Baziz, M., Lesbegueries, J., Gaio, M., 2007.

Une approche d'extraction et de recherche

d’information spatiale dans les documents textuels –

évaluation, In Proceeding of Conférence en Recherche

d’Information et Applications (CORIA), Saint-Etienne

France, pp. 53-64.

Sanderson, M., Kohler, J., 2004. Analysing Geographic

Queries, In Proceedings of SIGIR the Workshop on

Geographic Information Retrieval, Sheffield UK, pp. 8-

10.

Vargas, R.N.P., Moura, M.F., Speranza, E.A., Rodriguez,

E., Rezende, S.O., 2012. Discovering the Spatial

coverage of the documents through the SpatialCIM

Methodology, AGILE'2012 International Conference

on Geographic Information Science, Avignon, April

24-27, pp. 181-186.

Xueqing, L., Yangqiu, S., Shixia, L., Haixun, W., 2012.

Automatic taxonomy construction from keywords, In

Proceedings of the 18th ACM SIGKDD international

conference on Knowledge discovery and data mining

(KDD '12), ACM, New York, NY, USA, pp. 1433-

1441.

Yangqiu, S., Shixia, L., Xueqing, L., Haixun, W., 2015.

Automatic Taxonomy Construction from Keywords via

Scalable Bayesian Rose Trees, IEEE Transactions on

Knowledge and Data Engineering, Vol. 27, Issue 1, pp.

1861-1874.

Geographical Queries Reformulation using Parallel FP-Growth for Spatial Taxonomies Building

381