Searching Information Tourism using Vector Space Model

Dede Wintana

, Sfenrianto

, Hikmatulloh

, Mugi Raharjo

, Jordy Lesmana Putra

, Dyah Ayu

Ambarsari

, Desi Dwi Jayanti

Master of Computer Science Postgraduate Program, STMIK Nusa Mandiri, Jakarta, Indonesia

Information System management Departemen, BINUS Graduate Program, Master of Information System Management,

Bina Nusantara University, Jakarta, Indonesia

Keywords:

Tourism, Searching, Vector Space Model, Waterfall, Information Retrival.

Abstract:

Development of tourism is directed at creating tourist destinations that are evenly distributed in an area, includ-

ing tourism in Sukabumi, West Java, Indonesia. This study conducts secondary data retrieval from the internet

by using keyword ”air terjun sukabumi“. Information search method is to use a Vector Space Model (VSM) to

see the level of similarity with the weighting term. By passing various processes such as tokensizing, ﬁltering,

stemming, tf and df, calculation of inverse document frequency and several other processes. The results shows

that the proposed algorithm is very suitable for processing the sample. Determination of document ranking

is done by several stages of the algorithm that produce some documents with the highest value. The study

can also prove the proposed algorithm can work well in the case of searching for waterfall data in Sukabumi

district.

1 INTRODUCTION

West Java has one district with promising potential

for the advancement of tourism, namely Sukabumi

Regency, which is located in the Java south. In the

Regional Long Term Development Plan (RPJPD) of

Sukabumi Regency in 2005-2025 it was stated that the

priority of tourism development was directed at the

creation of tourism destinations in Sukabumi as one of

the leading tourism destinations in West Java. Indone-

sian tourism competition was increasingly sharp, so

that demanding every region explore potential sources

power to sell, attract and be visited by tourists (Darsi-

harjo, 2016). Eco-tourism is a type or type of tourism

that makes natural resources an object that is one of

the selling power of Sukabumi district, coupled with

artiﬁcial resources. For some areas in West Java, eco-

agro-tourism has developed well, but there are still

many other areas that have the potential to develop

eco-agro-tourism for the advancement of the region

and the welfare of its people (Hasugian, 2003).

From a development perspective, ecotourism busi-

nesses should only be considered ’successful’ if local

communities have some control over them and if they

share the beneﬁts equitably emerge from ecotourism

activities (Scheyvens, 1999). One of the visited ar-

eas at Sukabumi is the existing Ciletuh geopark in the

southern region of Sukabumi. The Geopark is an earth

park that is included in a conservation area, which has

elements of geodiversity, biodiversity, and cultural di-

versity which has aspects in the ﬁeld of education

as knowledge geology on the uniqueness and diver-

sity of the earth’s heritage in managing the area as

tourism (Darsiharjo, 2016). It has also many diversity

of tourist attractions such as beaches and many water-

falls that can be visited and waterfalls in the Ciletuh

area (Hardiyono et al., 2015). There are some wa-

terfalls in Ciletuh, namely: Awang waterfall, Curung

Puncak Manik, Sodong waterfall, Cimarinjung water-

fall and much more.

Collaboration is very important for tourism mar-

keting that is sucses, purpose, and electronic com-

munication are a new opportunity to be an opportu-

nity to work between tourism suppliers (Palmer and

Mccole, 2009). The increasing number of tourists to

Sukabumi is one way to increase regional income, but

until now tourism development is still not evenly dis-

tributed, especially the many waterfalls, consequently

even distribution of tourism objects is not yet optimal

so scientiﬁc studies are needed on tourist attraction in

Sukabumi.

Many countries promote nature-based tourism in

Indonesia to promote the purpose of nature conser-

vation and income generation (Hearne and Salinas,

Wintana, D., Sfenrianto, ., Hikmatulloh, ., Raharjo, M., Putra, J., Ambarsari, D. and Jayanti, D.

Searching Information Tourism using Vector Space Model.

DOI: 10.5220/0009882102470251

In Proceedings of the 2nd International Conference on Applied Science, Engineering and Social Sciences (ICASESS 2019), pages 247-251

ISBN: 978-989-758-452-7

247

2002). For natural resource management planning

planners make schemes and approaches to anticipate

various biased information caused by these limita-

tions. The characteristics of coastal resources and

Ciletuh geopark are quite complex and require spe-

ciﬁc management policies in order to provide optimal

and sustainable life and livelihood (Wahyudin, 2011).

In this study conducted research from the Internet

by using the keyword ”air terjun sukabumi“ to deter-

mine the extent to which tourists know about tourist

objects in the region. It is expected that with using

a Vector Space Model (VSM) can be known about

which regions little tourists for help the Sukabumi

tourism department in promoting tourism objects.

2 RESEARCH FRAMEWORK

The research framework starts from data presentation

(see Figure 1). The data that will be used is sec-

ondary data taken from the internet related waterfall

in Sukabumi on the internet. The data will be pro-

cessed using vector space model (SVM) to calculate

the amount of value obtained based on a keyword.

Figure 1: Research Framework.

The VSM is the basic method used for text repre-

sentation. To represent the text of each feature term

(Ti) considered as a coordinate in n-dimensional vec-

tor space, the corresponding weight (W1, W2, W3,

... Wn) is considered to be the coordinate of value.

Then, it can be used to represent text (Houy, 2013)

(Langcai et al., 2017). It is a method to see the level

of proximity or similarity term by weighting the term.

Documents are viewed as a vector that has a magni-

tude (distance) and direction. In the VSM, a term is

represented by a dimension of vector space. The rel-

evance of a document to a query is based on the simi-

larity between the document vector and the query vec-

tor (Amin, 2013) stages in the Information Retrieval

covering several processes in documents.

Then, indexing process to get the weight of each

term in the document. Calculation of these weights

is done by calculating the Term Frequency (Tf) and

Document Frequency (Df) of each term contained in

the document collection. Next process are tokenizing,

ﬁltering, and stemming.

The tokenizing process is done by a mechanism

if the document on the corpus is found in a space,

so the term between spaces will be retrieved by the

system. Then the term is placed in the initial table.

Process results in the form of original terms (terms

that still have additions, inherent punctuation marks,

and numbers).

The ﬁltering process is done by a mechanism if the

terms in the initial table table are found punctuation,

capital letters, and numbers. Then the program will

remove (punctuation and numbers) and change cap-

ital letters to lowercase letters, then check the term

with stopwords. The process results in the form of

selected terms (without punctuation, without capital

letters, and not including stopwords).

The Stemming process is a program for eliminat-

ing process or how to remove the rewards found in the

ﬁltering term. The eliminating process is done by re-

moving the preﬁx, insertion, and sufﬁx. The result of

the process is frekuency table. Finally the calculation

process and ranking of documents are carried out.

3 RESULT AND DISCUSSION

3.1 Data Preparation

The research data was obtained from searches on

Google’s search engine by inputting the request

”air terjun sukabumi”. There are 8 (eight) docu-

ment sample titles (D1 - D8) search results, including:

[D1] Curug Sawer, Air Terjun Eksotis di Suk-

abumi

[D2] Air Terjun Cikaso, Wisata Alam Andalan

Kabupaten Sukabumi

[D3] Air Terjun Pareang, kemegahan surga

tersembunyi di Sukabumi

[D4] Mengintip Blue Curug Cikaso yang mem-

pesona di Sukabumi Selatan

[D5] Curug Mawi, tempat baru untuk berada di

Cibadak Sukabumi

[D6] Air Terjun Gerong Sukabumi Air Terjun

Instagramable

ICASESS 2019 - International Conference on Applied Science, Engineering and Social Science

248

[D7] Eksotisme Air Terjun Cimarinjung, Wisata

Dunia Jurrasic dalam gaya Sukabumi

[D8] Curug Caweni, Alam Perawan Sukabumi

3.2 Tokenizing

After data is available, the next step is tokenize pro-

cess. Figure 2 shows the results of the tokenize which

have to group each word from the document that has

been obtained.

Figure 2: Tokenizing Process.

3.3 Filtering

After tokenizing, the words that appear in the data are

known. It has been applied for a few words, then it

will enter the ﬁltering stage.

Figure 3 - Figure 4 shows the ﬁltering results from

the data.

Figure 3: Filtering Process.

Figure 4: Filtering Process (Extension).

3.4 Stemming

In the Filtering process is data that generates 34

words. While in this process of stemming is 31 words

(see Figure 5 - Figure 6).

Figure 5: Stemming Process.

Figure 6: Stemming Process (extension).

3.5 Calculation

The calculation process consists of tf-idf, document

weight (W), distance Q - D, the document calculation

of DOT, and similarity. In the VSM approach, cal-

culations can be found based on the term frequency

Searching Information Tourism using Vector Space Model

249

and inverse document frequency (t f − id f ) based on

the log Number of Documents (D) / many documents

(df). For example ”Air” data, It is known: D =

8 ([D1], [D2], [D3], [D4], [D5], [D6], [D7], and

[D 8]); df (water) = 2; then t f − id f = log(8/2) =

0.6021. Then to calculate document weights using

equations w f (td) = t f (td)∗id f t. Based on document

”air” data in document 1 [D1], then t f (td) =;id f =

0.602.T husw f (td) = 1 ∗ 0.6021 = 0.6021.

To calculate the distance query (Q) to document

distance (D) using the equation : Sqrt(

∑

j=1

DOT product calculations use the equation:

∑

(Q ∗

) =

∑

i j

. The result of Q− D and DOT Prod-

uct from documents can be seen in Figure 7 and Fig-

ure 8.

Figure 7: Document Calculation of Distance.

Figure 8: Document Calculation of DOT product.

The next process is to calculate the cosine an-

gle between the vector Query with each document

by the formula: Consine θ D

= (Q ∗ D)/(|Q| ∗

|). The results of each document are Cosinus

[D1] = ((1.53) * (1,648809)) / (0,262778); Cosinus

[D2] = ((1.53) * (1,648809)) / (0,665157); Cosinus

[D3] = ((1.53) * (1,80618)) / (0); Cosinus [D4] =

((1.53) * (1,903881)) / (0); Cosinus [D5] = ((1.53)

* (2,019371)) / (0,665157); Cosinus [D6] = ((1.53)

* (1,648809)) / (0,262778); Cosinus [D7] = ((1.53)

* (1.676064)) / (0); and Cosinus [D8] = ((1.53) *

(1,411956)) / (0).

3.6 Document Ranking

From the results of the Vector Space Model (VSM)

analysis, the ranking results for the top four of the 8

documents that have been calculated are Ranking 1:

[D2] with a value of 0.262819; Ranking 2: [D5] with

a value of 0.21459; Ranking 3: [D1] with a value

of 0.10383; and Ranking 4: [D.6] with a value of

0.10382. So the documents that are most relevant to

the keywords ”air terjun sukabumi” is the document

4 or D4 = Mengintip Blue Curug Cikaso yang mem-

pesona di Sukabumi Selatan.

4 CONCLUSION

The study results show that the proposed algorithm

can function quite well in processing a sample of

8 (eight) documents. Determination of document

ranking is done by several stages of the VSM (Vec-

tor Space Model) algorithm that produces 4 (four)

large documents with the highest value, and the value

0.262819 becomes the highest value. It can prove that

the proposed algorithm can work well in the case of

search for keyword ”air terjun sukabumi”.

The development of this research still has to be

done, because in this study the data taken is still in

the scope of the Sukabumi district, so that more op-

timized results can then be taken in further develop-

ment within the scope of a wider waterfall data. In

further research can also be developed by trying sev-

eral algorithms that are more optimal in this study

case 6.

REFERENCES

Amin, F. (2013). Sistem temu kembali informasi dengan

pemeringkatan metode vector space model. 2(2):122–

129.

Darsiharjo, D. (2016). Pengembangan geopark ciletuh

berbasis partisipasi masyarakat sebagai kawasan ge-

owisata di kabupaten sukabumi. 13(1).

Hardiyono, A., Syafri, I., Rosana, M. F., Yuningsih, E. Y.,

and Andriany, S. S. (2015). Potensi geowisata di

kawasan teluk ciletuh, sukabumi, jawa barat. 13(2).

ICASESS 2019 - International Conference on Applied Science, Engineering and Social Science

250

Hasugian, J. (2003). Penggunaan bahasa alamiah dan kosa

kata terkontrol dalam sistem temu kembali informasi

berbasis teks.

Hearne, R. R. and Salinas, Z. M. (2002). The use of choice

experiments in the analysis of tourist preferences for

ecotourism development in costa rica. 64(2):153–163.

Houy, L. M. (2013). A generalized framework for ontology-

based information retrieval. pages 165–169.

Langcai, C., Zhihui, L., and Yuanfang, L. (2017). Research

of text clustering based on improved vsm by tf under

the framework of mahout. pages 6597–6600.

Palmer, A. and Mccole, P. (2009). International journal of

contemporary hospitality management emerald article

: The role of electronic commerce in creating virtual

tourism destination marketing organisations the role

of electronic commerce in creating virtual tourism

destination marketing. 12:198–204.

Scheyvens, R. (1999). Ecotourism and the empowerment of

local communities. 20:245–249.

Wahyudin, Y. (2011). Karakteristik sumberdaya pesisir

dan laut kawasan teluk palabuhanratu , kabupaten suk-

abumi , jawa barat. 1:19–32.

Searching Information Tourism using Vector Space Model

251