The "News Web Easy'' news service as a resource for teaching and ...

Proceedings of The 2nd Workshop on Natural Language Processing Techniques for Educational Applications, pages 73–81,Beijing, China, July 31, 2015. c©2015 Association for Computational Linguistics and Asian Federation of Natural Language Processing

The “News Web Easy” news service as a resource for teaching andlearning Japanese: An assessment of the comprehension difficulty of

Japanese sentence-end expressions

Hideki Tanaka, Tadashi Kumano and Isao GotoScience and Technology Research Labs. of NHK

1-10-11, Kinuta, Setagaya, Tokyo, Japan{tanaka.h-ja,kumano.t-eq,goto.i-es}@nhk.or.jp

Abstract

Japan’s public broadcasting corporation,NHK, launched “News Web Easy” inApril 2012 1. It provides users withfive simplified news scripts (easy Japanesenews) on a daily basis. This web ser-vice provides users with five daily sim-plified news scripts of “easy” Japanesenews. Since its inception, this servicehas been favorably received both in Japanand overseas. Users particularly appreci-ate its value as a Japanese learning andteaching resource. In this paper, we dis-cuss this service and its possible contri-bution to language education. We focuson difficulty levels of sentence-end expres-sions, compiled from the news, that cre-ate ambiguity and problems when rewrit-ing news items. These are analyzed andcompared within regular news and NewsWeb Easy, and their difficulty is assessedbased on Japanese learners’ reading com-prehension levels. Our results revealedthat current rewriting of sentence-end ex-pressions in News Web Easy is appropri-ate. We further identified features of theseexpressions that contribute to difficulty incomprehension.

1 Introduction

The convergence of TV and internet has enabledthe creation of new services that allow users toovercome various temporal and spatial constraints(Hamada, 2013; Fu et al., 2006). It may evenprove possible to effectively re-purpose contentacross different media. In this paper, we describeone such example: the application of TV newsscripts for language teaching and learning on theinternet.

1http://www3.nhk.or.jp/news/easy

Japan’s public broadcaster, NHK, launchedthe News Web Easy web service in April2012 (Tanaka et al., 2013). This site providesusers with five daily simplified news scripts ofeasy Japanese news. Its purpose is to provide dailynews to the foreign population in Japan, which hassteadily increased to currently over two million. Itwould, of course, be preferable to provide newsto these residents in their native languages. How-ever, Japan’s foreign population is so diverse thatit would be virtually impossible to provide broad-casts in all of the expatriates’ languages. NHKdecided to tackle this issue by providing broad-casting services in simplified Japanese tailored tothe language comprehension levels of foreign res-idents. Surveys among foreign residents have con-firmed that a demand exists for broadcasts in easyJapanese (Yonekura, 2012).

News Web Easy’s targeted audience in Japancomprises foreign residents learning Japanese as asecond language 2 who are already fairly fluent inconversational Japanese, but who want to learn toread news articles and newspapers. Their Japaneseis thus at a pre-intermediate level.

The easy Japanese news comprises regular newsthat is jointly rewritten by a Japanese languageinstructor, with special training in easy Japanese,and a reporter. They closely adhere to the basicvocabulary and sentence patterns listed in the testguidelines of the Japanese-Language ProficiencyTest (JLPT) (The Japan Foundation and Japan Ed-ucational Exchange and Services, 2002).

The test measures learners’ Japanese profi-ciency at four levels ranging from level 4 (elemen-tary) to level 1 (the most advanced) 3. The vocab-ulary lists and sentence patterns in the test guide-lines are graded, and the instructors can consult

2For the sake of brevity, in this paper we use the terms“foreigners” or “foreign residents” to signify foreign resi-dents learning Japanese as a second language.

3The test has now been revised to cover five levels rangingfrom N5 (elementary) to N1 (the most advanced).

73

these to verify the level of difficulty. News WebEasy employs vocabularies and sentence patternsranked at levels 3 and 4.

NHK has received favorable comments aboutNews Web Easy from foreign residents in Japanas well as from people overseas who appreciatethis service for learning Japanese. Japanese lan-guage instructors also regard News Web Easy asa valuable educational resource. We believe thatthis favorable reaction is the result of the languagelevel being set to an educational standard appropri-ate for Japanese. Moreover, the News Web Easyinterface is similar to that adopted in language tu-toring systems.

In this paper, we outline the features of NewsWeb Easy and discuss its impact on languagelearning and education. We focus analytically onsentence patterns (sentences-end expressions) innews scripts that are not adequately covered inthe JLPT test guidelines. Our analysis was basedon an extensive collection of these materials com-piled from a corpus of regular news texts and easyJapanese texts. We present an assessment of thedifficulty levels of these expressions according toforeigners’ reading comprehension levels ascer-tained from online questionnaires. Last, we con-sider the possibility of extending News Web Easyas a learning and teaching resource for native-levelJapanese used within regular news scripts.

2 News Web Easy and language teachingand learning

In this section, we explain the features of NewsWeb Easy and discuss its impacts on Japaneseteaching and learning.

2.1 Features of Japanese in News Web Easy

Target level

We were involved in the production of News WebEasy content. Our aim was to ensure that while thenews texts were easy to understand, they were asnatural as possible. After many trials conductedby NHK the pre-intermediate level was found tobe the lowest level necessary for achieving theseaims. This level was consequently set by NHKas the target for Japanese rewriting. It approxi-mately corresponds to the proficiency level rankedbetween levels 3 and 2 of the old JLPT, and be-tween levels N3 and N2 of the new JLPT.

RewritersFor the production of News Web Easy, five regu-lar news texts are chosen daily and rewritten by anews reporter and a Japanese instructor who per-form different tasks. While the reporter stream-lines the news texts and retains only the core infor-mation, the instructor simplifies difficult expres-sions.

Rewriting policiesWhen rewriting news articles, reporters andJapanese instructors confine themselves as muchas possible to elementary vocabulary and sentencepatterns. As noted above, rewriters use the JLPTtest guidelines to check words and sentence pat-terns. An editor is specially assigned for this pur-pose to News Web Easy.

It is noteworthy that reporters and instructorsalso use terms that are not listed. These includetechnical terms, proper nouns, and terms that fre-quently appear in news articles but are difficult tosimplify.

2.2 Features of News Web Easy interface

In addition to the above-mentioned measures usedto simplify Japanese, News Web Easy has severalreading support functions, described below andshown in Figure 1.

Furigana (ruby) charactersJapanese text is a combination of Chinese charac-ters (kanji), two types of Japanese phonetic sym-bols (hiragana and katakana), Latin characters(romaji), and numbers.

Kanji characters are notoriously difficult tomaster because there are so many of them and alsobecause the same characters can be read in differ-ent ways depending on the context. Foreign res-idents often find themselves unable to understandthe meaning of words written in kanji.

To assist them, very small hiragana characters,called furigana, are offered above all kanji char-acters in News Web Easy to indicate the pronun-ciation. This enhances the ability of foreign read-ers to understand the meanings of Japanese words,even if they are unable to read kanji.

GlossariesThe basic approach adopted by News Web Easy isto write simple Japanese using elementary vocab-ulary. However, it is not possible to simplify thevocabulary of all difficult terms.

74

All  Kanji  characters  have  Furigana    (phone6c  characterｓ)  

Underlined  words  have  dic6onary  explana6on  

Blue  words  are  place  names  Magenta  words  are  person’s  names

Text  is  read  out  by    speech  synthesizer    

Link  to  the  original  news    

Figure 1: Screen shot of News Web Easy

News Web Easy resolves this issue by provid-ing glossaries to explain difficult terminology. Onthe News Web Easy site, a glossary entry can beaccessed by simply positioning the cursor overa word. A pop-up explaining the term is thendisplayed. A dictionary for Japanese elementaryschool students was used to provide the glossaryentries.

Proper nounsProper nouns, not included within preexistingglossaries, inevitably appear in news articles. Onthe News Web Easy pages, different kinds ofproper nouns are highlighted in different colors tocapture the readers’ attention. The reader may notknow exactly what the terms mean, but at least thisfeature enables them to differentiate between thenames of people, places, and organizations.

Text to speechSome foreigners have difficulty reading Japanese,but are perfectly capable of understanding the textif it is read out to them. News Web Easy features atext-to-synthesized voice function to facilitate thismode of understanding.

Links to original newsBecause News Web Easy reporters usually con-dense information from the original news item,full details are available through a link providedto the source web page.

2.3 Teaching and learningThe content produced by broadcasters is quite of-ten used for language learning and teaching. Ac-cordingly, it is important that News Web Easy con-tributes to this educational purpose as long as themain goal of providing news to foreign residentsis not hampered. In this section, we discuss NewsWeb Easy’s contribution to Japanese teaching andlearning.

Contribution of Japanese textsNews Web Easy essentially delivers “authentic”documents at a pre-intermediate level in naturalJapanese. These texts are incorporated withinautomatic tutoring (learning) systems such asReader-Specific Lexical Practice for ImprovedReading Comprehension (REAP) (Brown and Es-kenazi, 2004). Such documents attract keen inter-est among educators, although texts in languagesother than English and French are rare (Uitden-bogerd, 2014). Thus, the simplified texts featuredin News Web Easy are all the more valuable.

Contribution of the interfaceThe interface of News Web Easy offers readingsupport, as elaborated in section 2.2. Readingsupport is often used in language tutoring sys-tems. For example, the Automatic Text Adapta-tion Tool (Burstein et al., 2007) automatically addsvocabulary support, automatic text reading by aspeech synthesizer, summary text, and a transla-

75

tion of the original text as “marginal notes.” Thereading support functions in News Web Easy can,therefore, be invaluable by providing simplifiedtexts for pre-intermediate level Japanese learners.

The interface could also facilitate understandingof regular news for foreigners. As noted above,News Web Easy provides links to the originalnews stories so readers can compare both versionsof the text. The scaffolding effect of providingsimplified text for reading original text is widelyrecognized (Burstein et al., 2007; Eskenazi et al.,2013; Petersen and Ostendorf, 2007). Thus, thiscomparative reading should contribute to the com-prehension of regular news.

2.4 Language-level issues

As noted in section 2.1, the News Web Easyrewriters currently use the JLPT test guidelines tocheck the language levels of words and sentencepatterns in news items.

Because Japanese lessons typically start withthe use of daily conversations, words and sentencepatterns specific to news texts are often lacking inthe JLPT test guidelines.

The rewriters have to judge for themselves thedifficulty levels of words and phrases that aremissing from the list. This could result in in-consistency in the language level of the simplifiedtexts. Therefore, the content of the JLPT guide-lines needs to be extended. As a first step towardthis, we decided to focus on sentence patterns thatwere not included in the guidelines.

3 Analysis of sentence-end expressions

The sentence patterns in the JLPT test guidelinestakes the form of a word sequence in the final po-sitions of a sentence. We refer to this as sentence-end expressions. In the next section, we will definethese and explain the features.

3.1 Features of Japanese sentence-endexpressions

Japanese is a subject-object-verb (SOV) type oflanguage in which predicates are positioned atthe end position of a sentence. Japanese predi-cates usually contain one content word followedby some function words. Content words are typ-ically verbs, nouns, and adjectives, and functionwords are auxiliary verbs, particles, formal nouns,and delexical (formal) verbs.

In this paper, we use the term sentence-endexpressions (SEEs) to signify the function wordsequence. SEEs add tense, polarity, voice, andmodality to a sentence which we refer to as func-tional information, or simply as function. Suchfunctions play an important role in deciding themeaning of a sentence.

SEEs may have more than one function linedup at the sentence end positions. We refer to suchlined up functions as the function sequence (FS).An SEE, therefore, has a FS whose length is atleast one 4 .

Because Japanese is an SOV type of language,SEEs may become quite long when the “O” is inan embedded sentence, as in S(SOV)V. Let us con-sider a sentence with a single function of probabil-ity:

X 社は 来年の 利益を 3 倍に するかもしれない (probability)．Xsha wa rainen no rieki wo 3bai ni surukamosirenai (probability).(X Inc. may (probability) triple theirprofit next year.)

This may be embedded in a sentence that endswithということです (toiukotodesu) (people say),which has a hearsay function, as in:

X 社は 来年の 利益を 3 倍にす る かもしれない (probability)ということです (hearsay)．Xsha wa rainen no rieki wo 3baini suru kamosirenai (probability)toiukotodesu (hearsay).(People say (hearsay) that X Inc.may (probability) triple their profit nextyear.),

The English predicates in the above examplesoccupy different positions and do not have linedup functions. However, the Japanese predicates(SEEs) of both the main and subordinate clausesare linked to form a long SEE with the followingFS: probability + hearsay (length 2). This com-plex structure is common in long Japanese SEEsand can be difficult for learners of the languageto understand. We, therefore, consider SEE re-wording to be essential for reducing the languagedifficulty level. We decided to extensively com-pile SEEs from regular news and News Web Easyto evaluate their difficulty for foreigners’ compre-hension.

4We consider the number of functions in FS as the length.

76

Regular Easy TotalSentence Count 3,937,214 20,616 -SEEs 477 775 1,063Meaningful SEEs - - 841

Table 1: Corpus size and SEE counts

3.2 Compilation of SEEs

We morphologically analyzed our corpus of reg-ular news scripts covering a 16-year period andsearched for SEEs. Our corpus contained aboutfour million sentences. We only selected those thatappeared over 100 times, resulting in a total of 466SEE sentence types. Although our selection wasrestricted to the above frequency threshold, it stillcovered 98% of the total occurrence of all SEEs.Considering the corpus size, we found that SEEvariation in the regular news was relatively lim-ited.

We also extracted SEEs from our corpus ofNews Web Easy scripts, collected over a two-yearperiod. This corpus contained about 20,000 sen-tences from which we obtained 755 SEE types.The total number of SEE types collected fromboth corpora was thus 1,063. We then excludedSEEs with a plain statement, that is, SEEs thatdid not contain any meaningful functional infor-mation. This yielded 841 SEE types. Table 1shows the corpus size and SEE counts.

3.3 Functions specific to news scripts

Before assigning a FS to each of the 841 SEEs,we first checked the SEEs and functions in a lead-ing Japanese grammar textbook (Nihongo KizyutuBunpo Kenkyukai, 2010). We found that someSEEs did not appear in that textbook and thus rep-resented new functions that we termed objectivityand perception groups.

ObjectivityTwo expressions—mono-da and koto-da—fellwithin this category. The formal noun, mono, haslittle meaning and simply refers to things in gen-eral. Another formal noun, koto, refers to gen-eral events. These terms are often added to sim-ple factual statements in news stories, as in irei(exceptional)-no koto-da. Although it is possibleto simply say irei-da, the addition of koto adds for-mality to the sentence. We believe this reflects thejournalistic tendency of describing events as ob-jectively as possible. We, therefore, termed this an

objectivity function.

Perception groupVerbs such as mieru (seem), kiku (hear), and omou(think) entail the modality of how the speaker rec-ognizes an statement’s proposition. We thus re-ferred to this modality as perception. We iden-tified several SEEs that varied in objectivity andcontained perceptions of third parties. Table 2presents a list of SEEs with the perception groupfunction.

The first expression, to-miteiru, comprises thecontent verb miru (see). The second expression,to-mirareteiru, is the passive version. Because,Japanese passive forms are often used without anagent (subject in a positive sentence), the personwho does the seeing is not specified in this case.This lack of specification increases the level of ab-straction of the sentence and adds objectivity.

The third expression, to-shiteiru, entails adelexical verb, suru (do) that ambiguously refersto miru (see), iu (say), and omou (think). Thisambiguity further increases the level of abstractionand objectivity of the sentence.

The last expression, to-sareteiru, is the pas-sivized version of to-shiteiru that we consider tohave the highest level of ambiguity and objectiv-ity.

Table 3 shows a list of all the functions used inthis study. These are divided into functions of syn-tax, common modalities, and regular news specificmodalities.

3.4 FS assignment to SEEsTo assign a FS to each of the 841 SEEs (describedin section 3.2), we first compiled a set of regu-lar expressions that linked function words to unitsbearing a single function. We then applied theseregular expressions to the 841 SEEs and assigneda FS to each SEE.

Each SEE with a FS had a number of occurrencecounts for each news type: regular and News WebEasy. We used these numbers to determine the as-sociation between FS and news type. An odds ra-tio was used to estimate the association:

O =p

1− p

1− q

q, (1)

where p is the relative frequency of a given FS innormal news and q is the relative frequency of theFS in News Web Easy. A FS whose odds ratio wasgreater than or equal to 1 was considered to have

77

SEE Function Explanation Objectivityto miteiru percept. see lowto mirareteiru pas.-percept. be seen middleto shiteiru amb.-percept. do (see, say or think) middleto sareteiru pas.-amb-percept. be done (seen, said, or thought) high

amb. = ambiguous, pas. = passive, percept. = perception

Table 2: SEEs bearing perception group function

Syntax causative passive aspect give-get changeexample parallel nominalization noun

Modality hope need order question will(Common) selection prohibition invitation guess probability

reason explanation change-guess(News specific) percept.group objectivity hearsay

Table 3: Functions assigned to SEEs

Length 1 2 3 4Regular 9(0.14) 35 (0.54) 21 (0.32) 0 (0)

Easy 13 (0.16) 43 (0.51) 22 (0.27) 3 (0.04)

Table 4: The distribution of FS types

an association with normal news; otherwise it hadan association with News Web Easy 5 .

Table 4 shows the number and relative fre-quency of FS types categorized by length andnews type. The numbers for both news typespeaked with the FS length of 2 and showed a sim-ilar distribution.

We calculated the relative frequency distribu-tion of FSs using the same categories as in Table4.The results are shown in Figure 2.

Because FS may have occurrence counts in bothnews types, we calculated the average relative fre-quency for each one. We found that FSs assciatedwith News Web Easy had a high frequency con-centration at length 1, while FSs associated withregular news peaked at length 2. We thereforeconcluded that SEEs with a single function werepreferred in easy Japanese news.

Next, we compared the unique single functionsthat appeared specifically in each news type . Wecollected these functions from FSs of length 1 andthe final functions in FSs of length 3 (see Table 4).Table5 summarizes these results. A sharp contrastis evident between the two types. Those functions

5We, therefore, considered the function sequence to beassociated with regular news if p was greater than or equal toq; otherwise, it was considered to be associated with NewsWeb Easy.

Easy

Regular

0.0

0.1

0.2

0.3

1 2 3 4Length of function sequence

Ave

rage

rela

tive

frequ

ency

of R

egul

ar a

nd E

asy

Figure 2: Average relative Frequency

used exclusively in regular news were all functionsin the perception group. Those used exclusively inNews Web Easy were syntactic types and modali-ties commonly used in daily conversation.

4 Evaluation of SEE difficulty forforeigners

4.1 Measure of difficultyThe difficulty levels of SEEs for foreign studentswere determined based on Japanese proficiencylevels. This was measured according to the newJLPT version, using the lowest grade required toread and understand SEEs. Since the new JLPThas five grades, ranging from N1 (the most ad-vanced) to N5 (elementary), we attached numbers

78

Length Easy Regular1 give-get amb.-percept.

order pas.-percept.probability pas.-amb.-percept.prohibition

3 (final) order amb.-percept.explanation pas.-percept.reason

Table 5: Single functions unique to each news type

Grade Number Grade NumberN5 1 N2 4N4 2 N1 5N3 3 above N1 6

Table 6: JLPT levels and numbers for selection

ranging from 1 to 5 to them, with 5 indicating themost difficult SEEs and 1 the easiest. The number6 was designated to SEEs that were difficult, evenfor N1-grade students. Table 6 presents the JLPTgrades and numbers for the selection.

4.2 Selection of SEEs

We aimed to evaluate the difficulty levels of the841 SEEs for foreign students learning Japaneseand to analyze the factors governing these diffi-culty levels. The total number of SEEs (841) wastoo high to evaluate individually. Moreover, theword types for building SEEs were too diverse forthe extraction of just a few factors.

We, therefore, decided to first sample FSs andthen select SEEs bearing the sampled FSs. Thenumber of FS types was 146 (Table 4) and that offunction types was only 28 (Table 3) which wouldresult in a highly tractable analysis of FSs. Ac-cordingly, we selected SEEs based on the follow-ing assumptions and procedures.

(A1) Any FS belonging to the same cell in Table 4would have the same difficulty level 6 .Based on this assumption, we randomly sam-pled 13 FSs from the four cells in Table 4 oflength 1 and 3, and for both news types. Thisresulted in a sample of 52 FSs.

(A2) Any SEE belonging to the same FS had thesame difficulty level. Based on this assump-tion, we selected the most frequently occur-

6In other words, the difficulty level of FS only depends onthe news type and the length of FS.

Target level (3.5)

Easy

Regular

3.0

3.2

3.4

3.6

3.8

1 3Length of function sequence

Ave

rage

sco

re

Figure 3: Score difference for the SEEs

ring SEEs from each of the 52 FSs sampledin (A1) which also yielded 52 SEEs.

Because the regular news cell with a length of 1had nine functions, we sampled 13 SEEs, allow-ing for FS duplication. It should be noted that thedifficulty level of each sampled SEE and its FSwas considered to be equal, because FS and SEEcorresponded on a one-to-one basis.

4.3 Subjects and questionsWe believed that foreign students—especially atthe elementary level—would find it very difficultto respond to questions about their comprehensionlevels of SEE as they would need an in depth un-derstanding of the functions to do so. We, there-fore, asked Japanese instructors, and not the stu-dents, directly, to evaluate the difficulty levels ofthe SEEs.

Each questionnaire for the 52 SEEs consistedof the following parts: the SEE in question; func-tional features; and examples of usage in sen-tences. We sent questionnaires to 500 Japanese in-structors through Internet. They specified the dif-ficulty number in Table 6 for each of the 52 SEEs.In total, 390 effective responses were returned tous.

4.4 Results and discussionBased on the responses of the 390 instructors, Fig-ure 3 shows the average numbers (scores) calcu-lated for the difficulty of SEEs and FSs obtainedfor the four cells (see Table4). The x-axis showsthe FS length and the y-axis shows the average

79

Function News type Scorepas.-percept. regular 3.800pas.-amb.-percept. regular 3.791percept. regular 3.736noun regular 3.731amb.-percept. regular 3.701objectivity regular 3.513hearsay regular 3.485causative regular 3.369passive regular 3.354probability easy 3.346

Table 7: The top 10 single functions according todifficulty levels

scores. The solid line indicates score changes forNews Web Easy and the dotted line shows thosefor regular news. It should be noted that we setour target level of Japanese between N3 and N2 ofthe current JLPT. The target threshold score was3.5.

The graph shows that the difficulty level ofSEEs with a FS length of 1 from the News WebEasy cell was under the target threshold, whilethose in other cells were above this level.

Effect of news type and FS length

From that graph, it is evident that for both thelengths, SEEs obtained from News Web Easy wereeasier to comprehend than those obtained for reg-ular news. We may conclude that the rewritingof the SEEs evidently reduced difficulty in under-standing.

SEEs of a FS length of 3 were more difficultthan those of a length of 1 for both types of news.As can be seen from the difference in the lines’gradients, the increase in difficulty associated withan increase in FS length was more apparent forSEEs from News Web Easy 7 . Although singlefunctions used in News Web Easy evidenced lowdifficulty levels, these levels rose rapidly whenthey were lined up. Length of SEE is obviouslyone of the factors that affect the difficulty level.

Effect of functions

To confirm individual differences in FSs found inregular news and News Web Easy, we focused onthe FSs with a length of 1 and arranged them,score-wise, in descending order (Table 7).

7The two-way analysis of variance test revealed that thedifference was statistically significant.

It is evident from Table 7 that the nine singlefunctions that appeared in regular news (see Ta-ble 4) occupied the top nine positions. In partic-ular, the perception group functions were consid-ered the most difficult. These were the ones thatonly appeared in regular news (c.f. Table 5). Wecan, therefore, conclude that many Japanese in-structors consider the elimination of these func-tions in easy Japanese news to be an appropriateapproach for maintaining the difficulty level belowthe pre-intermediate level.

5 Conclusions and future work

We were involved in NHK’s web service initia-tive, News Web Easy. This initiative aims to de-liver news in simplified Japanese to foreign resi-dents learning Japanese as a second language. Aswe reported, the service has been welcomed as aJapanese teaching and learning resource. For thisstudy, we analyzed features of News Web Easythat contribute to learning the language.

We focused on SEEs occurring in news mostof which are not listed in JLPT test guidelines.We compiled an extensive collection of SEEs fromregular news texts and News Web Easy and identi-fied differences in SEE usage within regular newsand News Web Easy. Consequently, we found asharp contrast in terms of grammatical functions.We then examined the difficulty levels of these ex-pressions for foreign students learning Japanesebased on a wide-scale evaluation by Japanese in-structors. Our results revealed that the currentrewriting of SEEs is appropriate. Moreover, weidentified features of these expressions that con-tribute to the difficulty factor.

A future challenge entails extending News WebEasy to make it a useful resource for those whowish to follow regular news that is written innative-level Japanese. Because News Web Easyfacilitates comparative reading of both normal andeasy Japanese, it offers such an opportunity tosome extent. To further enhance this function,we believe that the findings of the present studywill be valuable. The difficult SEEs that we foundwere appropriately reworded into simpler expres-sions and became unnoticeable in the simplifiedtexts. If we can explicitly provide feedback aboutsuch information to News Web Easy users, theywill be able to learn native-level Japanese more ef-ficiently. Creating such an interface is, therefore,part of our future plans.

80

ReferencesJonathan Brown and Maxine Eskenazi. 2004. Re-

trieval of authentic documents for reader-specificlexical practice. In Proceedings of InSTILL/ICALLSymposium.

Jill Burstein, Jane Shore, John Sabatini, Yong-WonLee, and Matthew Ventura. 2007. The automatedtext adaptation tool. In Proceedings of HumanLanguage Technologies: The Annual Conference ofthe North American Chapter of the Association forComputational Linguistics (NAACL-HLT), pages 3–4, April.

Maxine Eskenazi, Yibin Lin, and Oscar Saz. 2013.Tools for non-native readers: the case for translationand simplification. In Proceedings of the Workshopon Natural Language Processing for Improving Tex-tual Accessibility, pages 20–28, June.

Hsin Chia Fu, Yeong Y. Xu, and C.L. Tseng. 2006.Generation of multimedia TV news contents forWWW. In Proceedings of the 15th InternationalConference on World Wide Web, pages 909–910.ACM, May.

Hiroyuki Hamada. 2013. Overview of the hybridcastsystem. Broadcast Technology, 51:1–8.

Nihongo Kizyutu Bunpo Kenkyukai, editor. 2010.Modern Japanese Syntax (Gendai Nihongo Bunpo)1-7. Kuroshio Shuppan. (in Japanese).

Sarah E. Petersen and Mari Ostendorf. 2007. Text sim-plification for language learners: A corpus analysis.In Proceedings of ISCA SLaTE2007, pages 69–72.

Hideki Tanaka, Hideya Mino, Shinji Ochi, and MotoyaShibata. 2013. News services in simplified Japaneseand its production support systems. In Proceedingsof the International Broadcasting Convention 2013.

The Japan Foundation and Japan Educational Ex-change and Services, editors. 2002. Japanese–Language Proficiency Test: Test Content Specifica-tions (Revised Edition). Bonjinsha. (in Japanese).

Alexandra Uitdenbogerd. 2014. Tools for supportinglanguage acquisition via extensive reading. In Work-shop Proceedings of the 22nd International Confer-ence on Computers in Education, pages 35–41.

Ritsu Yonekura. 2012. Information searchand media access of foreign residents in Japanin disaster period—telephone interviews on fournationalities—. The NHK Monthly Report onBroadcast Research, pages 62–75, August. (inJapanese).

81