TF-IDF TDF-IDF TDF-IDF Extracting Impression of Sightseeing Spots from Blogs for Supporting Selection of Spots to Visit in Travel Sat

Size: px

Start display at page:

Download "TF-IDF TDF-IDF TDF-IDF Extracting Impression of Sightseeing Spots from Blogs for Supporting Selection of Spots to Visit in Travel Sat"

えつみいしなみ
5 years ago
Views:

1 TF-IDF TDF-IDF TDF-IDF Extracting Impression of Sightseeing Spots from Blogs for Supporting Selection of Spots to Visit in Travel Satoshi Date, 1 Teruaki Kitasuka, 1 Tsuyoshi Itokawa 2 and Masayoshi Aritsugi 1 In this paper, we try to recommend spots to visit in travel using impression of sightseeing spots from blogs. We recommend spots having same impression as those the user visited and/or likes. We suppose impression can be expressed as nouns, adjectives or verbs in blogs. We first collect blogs concerning sightseeing spots. We then morphological analyze them and extract nouns, adjectives and verbs and assign to them TF-IDF scores calculated by a TDF-IDF based method. Lastly, we find candidate spots to visit that are supposed to be similar to favorite spots of the user by using Pearson product-moment correlation coefficients. For evaluating which part or a combination of parts of speech can work well for finding good spots, we compare seven combinations of nouns, adjectives and verbs. The results show that nouns themselves could work well in six out of 18 cases in the experiments. Also, they show there are some cases where each combination could work well. 1. TF-IDF Term Frequency-Inverce Document Frequency 1 Graduate School of Science and Technology, Kumamoto University 2 Kumamoto Prefectural College of Technology

2 Yu 1) GPS GPS GPS GPS GPS GPS GPS Yu GPS HITS HITS Hyperlink-Induced Topic Search ( ) Web ( ) Web Web Web Web Yu Yu 2) GPS Java PC 2.2 Pavel 3) flickr flickr flickr flickr Pavel 4)

3 5) TDF-IDF TDF-IDF Fig. 1 Process flow of proposal URL TF-IDF TDF-IDF Term s Document Frequency-Inverce Document Frequency i TDF-IDF (1) i TF TF TF TF (2) i TDF (2)

4 i i TF-IDF 2 TDF i TF = i TDF = n i Σ k n k (1) i (3) i IDF IDF i i IDF = log i i TDF-IDF = i TDF i IDF (4) (4) i TDF-IDF TDF-IDF TDF IDF TDF-IDF 3.3 TDF-IDF r (5) r(x, Y ) = Σ n i=1(x i X)(Y i Y ) Σ n i=1 (X i X) 2 Σ n i=1 (Y i Y ) 2 (5) X Y 2 i n X Y X i X i TDF-IDF (2) (3) Y i Y i TDF-IDF X X TDF-IDF Y Y TDF-IDF TDF-IDF Table 1 Sightseeing spots with their categories A I II III B IV V VI C VIIVIIIIX D X XI XII E XIII XIV XV F XVI USJ XVII XVIII

5 Table 2 Results of Impression words and sightseeing spots TDF-IDF 3 Table 3 The number of parts of speech s kinds USJ TDF-IDF TDF-IDF 6 1 4

6 2 TDF-IDF TDF-IDF TDF-IDF TDF-IDF TDF-IDF 5 TDF-IDF 4.4 TDF-IDF TDF-IDF TDF-IDF TDF IDF TDF-IDF TDF-IDF TDF-IDF TDF-IDF TDF-IDF TF-IDF TDF-IDF.

7 TF-IDF TDF-IDF ) Yu Zheng, Lizhu Zhang Xing Xie, Wei-Ying Ma Mining Interesting Locations and Travel Sequences from GPS Trajectories Proceedings of the 18th international conference on World wide web pp ) P-Tour:, Vol.45, No.12, pp ) Pavel Serdyukov, Vabessa Murdock, Roelof van Zwol Placing Flickr on a map Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval pp ) Vol.44, No.7, pp ) Vol.J82-D-I, No.1, pp

8 Table 4 Degree of similarities using nouns 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

9 Table 5 Degree of similarities using adjectives 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

10 Table 6 Degree of similarities using verbs 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

11 Table 7 Degree of similarities using nouns and adjectives 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

12 8 Table Degree of similarities using nouns and verbs 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

13 Table 9 Degree of similarities using adjectives and verbs 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

14 Table 10 Degree of similarities using nouns and adjectives and verbs 10 3 I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII I A II III IV B V VI VII C VIII IX X D XI XII XIII E XIV XV XVI F XVII XVIII

Step2 入門

Step2 入門 ii iii iv v vi NEC Corporation 1999 vii C O N T E N T S PART 1 PART 2 PART 3 viii PART 4 ix C O N T E N T S PART 5 x PART 6 xi C O N T E N T S PART 7 xii PART 8 PART 9 xiii C O N T E N T S xiv xv PART