Yahoo!search and Web Api Utilized Mashup Based E-leaning Content Search Engine for Mobile Learning

—Mashup based content search engine for mobile devices is proposed. Mashup technology is defined as search engine with plural different APIs. Mash-up has not only the plural APIs, but also the following specific features, (1) it enables classifications of the contents in concern by using web 2.0, (2) it may use API from the different sites, (3) it allows information retrievals from both sides of client and server, (4) it may search contents as an arbitrary structured hybrid content which is mixed content formed with the individual content from the different sites, (5) it enabling to utilize REST, RSS, Atom, etc. which are formed from XML conversions. The mash-up should be a flexible search engine for any purposes of content retrievals. The proposed search system allows 3D space display of search menus with these peculiarities on Android devices. The proposed search system featuring Yahoo!search BOSS and Web API is also applied for e-learning content retrievals. It is confirmed that the system can be used for search a variety of e-learning content in concern efficiently.


I. INTRODUCTION
E-learning content retrievals with mobile terminals have not been done easily due to the fact that display size is not good enough for search as well as there are a variety of contents types, video, document, etc.On the other hand, mashup technology [1], [2] supported web 2.0 [3] allows gathering information of e-learning contents efficiently and effectively.Meanwhile, visualization of web contents has been well developed.Therefore, e-learning content search engine can be improved in terms of efficiently and effectively based on the aforementioned mashup technology and visualization tools1.
Web API [4] based e-learning content search engine is developed with Yahoo!search BOSS (Build your Own Search Service) [5] which is called ELDOXEA: E-Learning Content Search Engine [6]- [10].Not only major keyword, but also minor keywords derived from the descriptions and keywords in the header information of URL which is hit by the first major keyword are used for search.The search engine proposed here is ELDOXEA for mobile devices.One of the 1 https://supportcenter.checkpoint.com/supportcenter/portal?eventSubmit_doGo viewsolutiondetails=&solutionid=sk30765 disadvantages of mobile devices is relatively small size of display.Therefore, in particular, representation method for search results has to be considered for fitting the search result contents efficiently and effectively.
There are many web visualization tools.In particular, UNIX software of Natto view [11] and mashup based search engine with www visualization of Flowser2 are sophisticated tools.With the reference to these software tools, the proposed e-learning content search engine is designed.E-learning content search engine with web API based approach is proposed in the following section.Bing and Google API are additionally featured.Furthermore, thesaurus based search results prioritized method is featured.Application software tool for the search engine is described followed by implementation and some experimental results.Then concluding remarks are described with some discussions.

A. Mashup Technology
Mashup is defined as a method for providing web services and applications based on combination of more than two web APIs.It allows improve usability by combining search results which are obtained from the different web APIs separately.Search results from web APIs, in general, in the form of XML or JSON [12].Therefore, it is easy to combine the search results into one.The following four web APIs are used for the proposed search engine,

B. Search Result Display
In order to display search results in a comprehensive manner, the following four candidate models, helix, star, star helix and star slide are trade-off.
One of important roles of the displaying search results is to show the priority of the results.From the top to bottom priority, "helix model" of displaying method allows priority representation from the top to the bottom of the display positions in a spiral order.It, however, is not appropriate displaying method because there are many types of the search results, document, URL, image, etc.Therefore, it is complicated and overlapped so much among the search result types.
Another model of the search result representation is "star model".This model is adopted by Flowser.The search results are classified by content types of nodes.From the top node to the node in concern, the length of rink between both shows priority.It, however, is not so easy to see because the nodes are located in the 3D domain of the display positions.
Star-Helix model combines the aforementioned two models, star and helix models.It, however, is still difficult to see the priority.Star-Slide model of search result display method is the most appropriate model as shown in Fig. 1.In the Fig. 1, star shows search result types while smile marks shows contents (URLs).From the top to bottom priority, "star-slide model" of displaying method allows priority representation from the top to the bottom of the display positions.It has to be displayed onto mobile device displays in a 3D CG representation.Away3D [9] is used for this purpose.There are a plenty of 3D CG representation software tools, OpenGL, ADOBE AIR, etc. under the programming language of Java, Objective-C, etc.In particular, Away3D is used on the ADOBE AIR (Stage3D) of Action Script as a library.The geometric relation between camera and object is shown in Figure 2 as well as rendering of object together with staging device.

C. LEDOXEA
LEDOXEA is mobile devices version of ELDOXEA: E-Learning Content Search Engine.After the activation of LEDOXEA, initial image of start-up is displayed onto mobile device display as shown in Fig. 3 (a).Immediately after this, key-in is available with screen keyboard as shown in Fig. 3 (b).The icon sizes are different by content type for representation of the search results in 3D space on the display as shown in Fig. 7.When users tap the title summary which is located the bottom of the display as shown in Fig. 8, then users may take a look at the search result of contents through browser or external applications.There is "back button" for getting back to the previous display image after the referencing the search result contents.Search content types are rotated when users swipe the content type icons.These rotation angles can be calculated as shown in Fig. 8.
Icon operations with swipe actions are illustrated in Fig. 9. Swipe action in the vertical direction allows search results candidate selections.Meanwhile, swipe action in the horizontal direction allows content type selections.These can be done through camera position changes in 3D space in the CG space of Away 3D.

A. LEDOXEA version 2
From the web site stability or availability points of view, the proposed search engine employed Bing API instead of Yahoo!Search.Display image layout is shown in Fig. 10.

B. Thesaurus Approach
In order to improve effectiveness of search engine, not only major keyword but also minor keywords can be used for e-learning content search.In addition to the dialog box, thesaurus search is available for the LEDOXEA version 2. Example of the keyword input image display is shown in Fig. 11 (a).There is the dialog box for thesaurus search which is situated just under the dialog box for the major keyword.Example of the image of thesaurus search results is shown in Fig. 11 (b).In this example, "Wikipedia" of thesaurus search engine is selected.Appropriate minor keywords are available as the thesaurus search results.Therefore, much appropriate elearning content search can be done with major and minor keywords.
Fig. 12 (a) shows the other thesaurus search result by keyword types abstract, history, word, footnote, reference, related site, other link are available for thesaurus search.Also, several languages are available for search by clicking the bottom button.Fig. 12 (b) shows an example of e-learning content search with the major keyword of "Java".When users key-in the minor keyword of "文法" (grammar in English) in the dialog box of thesaurus search engine, then they can get the much appropriate search results rather than LEDOXEA version 1 (without thesaurus engine).Fig. 13 (a) shows an example of search image when users key-in the major keyword "Java" and the minor keyword "grammar" for the thesaurus search engine while Fig. 13 (b) shows an example of search image when users key-in the major keyword "Java" and the minor keyword "question and answer" for the thesaurus search engine.Twenty of university students (information science department) participate the experiments.They use the LEDOXEA and then they evaluate the LEDOXEA.Their comments and suggestions are as follows, 1) It is comfortable to use the LEDOXEA because operability using swipe operations, smooth rendering, usability of Andoroid tablet terminals, WWW visualizations are excellent, 2) Search results can be seen simultaneously, 3) PDF data and YouTube of search results can be accessible from the LEDOXEA directly, 4) Appropriate e-learning contents can be retrieved efficiently with a single keyword.
5) Prioritizing the search results is excellent because the most appropriate e-learning content can be found easily, 6) Five e-learning contents can be seen simultaneously with a single keyword, 7) Still pictures as well as moving pictures are displayed their thumbnail, it is comprehensive.

IV. CONCLUSION
Mashup based content search engine with mobile terminals is proposed.Mashup technology is defined as search engine with plural different APIs.Mash-up has not only the plural APIs, but also the following specific features.it enables classifications of the contents in concern by using web 2.0, it may use API from the different sites, it allows information retrievals from both sides of client and server, it may search contents as an arbitrary structured hybrid contents which is mixed contents formed with the individual contents from the different sites, it enabling to utilize REST, RSS, Atom, etc. which are formed from XML conversions.
Although mashup allows content search which is same as portal, mashup has the aforementioned different features from portal.Therefore, mash-up is possible to create more flexible search engine for any purposes of content retrievals.The search system which is proposed here is that make it possible to control the graph in the 3D space display with these peculiarity on Android devices.The proposed search system is applied for e-learning content retrievals.It is confirmed that the system enables to search a variety of content in concern efficiently.
It is found that users can take a look at the search results from the top to the bottom priority of the search results by content type by type.Also, the highest priority of search result appears at the bottom of the image with hyperlink.In this example, the book on Java language specifically programming grammar is greatly featured is shown as search result.Users may input minor keywords on their own.Users also may use thesaurus search engine.It is totally up to users.

Fig. 2 .
Fig. 2. Geometric relation among 3D view port, camera, object and 3D area definition for rendering

Fig. 4 .
Fig. 4. Display image In this case, five content types and icons of search results of URLs appear in order of their priorities.Five content types

Fig. 5 .
Fig. 5. Content types for search Search results can be moved up and down when users swipe in the vertical direction as shown in Fig.6 (a).Therefore, search result content can be seen one by one.When users would like to see other types of search results of contents, they have to swipe the display in the horizontal direction as shown in Fig.6 (b).

Fig. 6 .
Fig. 6.Move up and down as well as ration with vertical swipe and horizontal swipe

Fig. 8 .
Fig. 8. Rotation angle of the five icons

Fig. 10 .
Fig. 10.Example of search result display with Bing API based LEDOXEA YouTube (moving pictures), Amazon site, ADOBE content (documents), and Image contents are available for search.These are common to the previous LEDOXEA featuring Yahoo!Search.Search results are aligned in the order of their priority.When users choose one of the icon of search result, then the hyperlink of the search result appears on the bottom of the screen.Also, users get the contents when they click the hyperlinked search result.This is referred to the version 1 of LEDOXEA hereafter.

Fig. 12 .
Fig. 12. Search keyword types selection for thesaurus search and example of e-learning content search with the major keyword of "Java" (a) minor keyword "grammar" (b) minor keyword "question and answer" www.ijarai.thesai.org

Fig. 15 .
Fig. 15.Example of linked retrieved results (Merged contents of Yahoo search results with the other retrieved contents)