Rodrigo Gomes de Oliveira, Yaji Sripada, Ehud Reiter
We describe an initial version of an algorithm for generating named references to locations of geographic scale. We base the algorithm design on evidence from corpora and experiments, which show that named entity usage is extremely frequent, even in less obvious scenes, and that names are normally used as the first focus on a global region. The current algorithm normally selects the Frames of Reference that humans also select, but it needs improvement to mix frames via a mereological mechanism.
{"title":"Designing an Algorithm for Generating Named Spatial References","authors":"Rodrigo Gomes de Oliveira, Yaji Sripada, Ehud Reiter","doi":"10.18653/v1/W15-4723","DOIUrl":"https://doi.org/10.18653/v1/W15-4723","url":null,"abstract":"We describe an initial version of an algorithm for generating named references to locations of geographic scale. We base the algorithm design on evidence from corpora and experiments, which show that named entity usage is extremely frequent, even in less obvious scenes, and that names are normally used as the first focus on a global region. The current algorithm normally selects the Frames of Reference that humans also select, but it needs improvement to mix frames via a mereological mechanism.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128703480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the present study, we investigate if speakers refer to moving entities in route directions (RDs) and how listeners evaluate these references. There is a general agreement that landmarks should be perceptually salient and stable objects. Animated movement attracts visual attention, making entities salient. We ask speakers to watch videos of crossroads and give RDs to listeners, who in turn have to choose a street on which to continue (Experiment 1) or choose the best instruction among three RDs (Experiment 2). Our results show that speakers mention moving entities, especially when their movement is informa- tive for the navigation task (Experiment 1). Listeners understand and use moving landmarks (Experiment 1), yet appreciate stable landmarks more (Experiment 2).
{"title":"Moving Targets: Human References to Unstable Landmarks","authors":"A. Baltaretu, E. Krahmer, A. Maes","doi":"10.18653/v1/W15-4706","DOIUrl":"https://doi.org/10.18653/v1/W15-4706","url":null,"abstract":"In the present study, we investigate if speakers refer to moving entities in route directions (RDs) and how listeners evaluate these references. There is a general agreement that landmarks should be perceptually salient and stable objects. Animated movement attracts visual attention, making entities salient. We ask speakers to watch videos of crossroads and give RDs to listeners, who in turn have to choose a street on which to continue (Experiment 1) or choose the best instruction among three RDs (Experiment 2). Our results show that speakers mention moving entities, especially when their movement is informa- tive for the navigation task (Experiment 1). Listeners understand and use moving landmarks (Experiment 1), yet appreciate stable landmarks more (Experiment 2).","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122926417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
JSrealB is an English and French text realizer written in JavaScript to ease its integration in web applications. The realization engine is mainly rule-based. Table driven rules are defined for inflection and algorithmic propagation rules, for agreements. It allows its user to build a variety of French and English expressions and sentences from a single specification to produce dynamic output depending on the content of a web page. Natural language generation can automate a significant part of textual production, only requiring a human to supply some important aspects and thus saving considerable time for producing consistent grammatically correct output. In recent years, tools such as SimpleNLG (Gatt and Reiter, 2009) facilitated text realization by a programmer provided they program their application in Java. This system was then extended with SimpleNLG-EnFr (Vaudry and Lapalme, 2013), a English-French version of SimpleNLG. Another approach to text realization is JSreal (Daoust and Lapalme, 2014), a French Web realizer written in JavaScript. This paper describes an attempt at combining the ideas of SimpleNLGEnFr and JSreal to produce a bilingual realizer for French and English from a single specification. JSrealB generates well-formed expressions and sentences. It can be used standalone for linguistic demonstrations or be integrated into complex text generation projects. But like JSreal, it is aimed at web developers, from taking care of morphology, declension and conjugation to creating well-formed texts. A web programmer who wishes to use JSrealB to produce flexible English and/or French textual or HTML output only needs to add two lines in the page: one for importing program and one for calling JSrealB loader to load the resources (i.e. lexicon and rules). The principles underlying JSrealB are similar to those of SimpleNLG: programming language instructions create data structures corresponding to the constituents of the sentence to be produced. Once the data structure (a tree) is built in memory, it is traversed to produce the list of tokens of the sentence. This data structure is built by function calls whose names are the same as the symbols usually used for classical syntax trees: for example, N to create a noun structure, NP for a Noun Phrase, V for a Verb, D for a determiner, S for a Sentence and so on. Features added to the structures using the dot notation can modify the values according to what is intended. JSrealBsyntactic representation is patterned after classical constituent grammar notations. For example, S(NP(D("a"),N("woman")).n("p"), VP("eat").t("ps")) is the JSrealB specification for The women ate. Plural is indicated with feature n("p") where n indicates number and p plural. The verb is conjugated to past tense indicated by the feature tense t and value ps. Agreement between NP and VP is performed automatically. French and English are languages whose structures are similar. Both languages use the same alphabet, they are bot
JSrealB是一个用JavaScript编写的英语和法语文本实现器，以简化其在web应用程序中的集成。实现引擎主要是基于规则的。表驱动规则被定义为拐点和算法传播规则，用于协议。它允许用户从单一规范中构建各种法语和英语表达式和句子，以根据网页内容产生动态输出。自然语言生成可以自动化文本生成的重要部分，只需要人类提供一些重要方面，从而节省大量时间来生成一致的语法正确的输出。近年来，SimpleNLG (Gatt和Reiter, 2009)等工具促进了程序员在Java中编写应用程序的文本实现。然后使用SimpleNLG- enfr (Vaudry and Lapalme, 2013)对该系统进行扩展，这是SimpleNLG的英法版本。另一种实现文本的方法是JSreal (Daoust and Lapalme, 2014)，一个用JavaScript编写的法语Web实现程序。本文描述了将SimpleNLGEnFr和JSreal的思想结合起来，从单一规范中生成法语和英语双语实现器的尝试。JSrealB生成格式良好的表达式和句子。它可以单独用于语言演示或集成到复杂的文本生成项目中。但像JSreal一样，它的目标用户是web开发人员，从处理词法、词形变化和词形变化到创建格式良好的文本。如果web程序员希望使用JSrealB生成灵活的英语和/或法语文本或HTML输出，只需要在页面中添加两行:一行用于导入程序，另一行用于调用JSrealB加载器来加载资源(即词典和规则)。JSrealB的基本原理与SimpleNLG类似:编程语言指令创建与要生成的句子成分相对应的数据结构。一旦在内存中构建了数据结构(树)，就遍历它以生成句子的标记列表。此数据结构由函数调用构建，其名称与经典语法树中通常使用的符号相同:例如，N用于创建名词结构，NP用于名词短语，V用于动词，D用于限定词，S用于句子，等等。使用点表示法添加到结构中的特性可以根据需要修改值。JSrealBsyntactic表示法是按照经典的组成语法表示法模式化的。例如，S(NP(D(“a”)，N(“woman”)). N(“p”)，VP(“eat”).t(“ps”))是女性吃的JSrealB规范。复数用特征n(“p”)表示，其中n表示数字，p表示复数。动词与过去时结合，由特征时态t和值ps表示。NP和VP之间自动执行一致。法语和英语是结构相似的语言。两种语言使用相同的字母表，它们都是融合语言，具有相似的词形系统，它们的词序遵循相同的基本主词客体范式(Shoebottom, 1996)。但结构上的差异确实存在:例如，这些语言之间形容词的位置不同，名词和代词的性别和数量一致规则也不同。这些差异必须在许多层面上加以考虑。首先，算法必须在短语或句子层面处理句法差异和一致(即特征传播)
{"title":"JSrealB: A Bilingual Text Realizer for Web Programming","authors":"Paul Molins, G. Lapalme","doi":"10.18653/v1/W15-4719","DOIUrl":"https://doi.org/10.18653/v1/W15-4719","url":null,"abstract":"JSrealB is an English and French text realizer written in JavaScript to ease its integration in web applications. The realization engine is mainly rule-based. Table driven rules are defined for inflection and algorithmic propagation rules, for agreements. It allows its user to build a variety of French and English expressions and sentences from a single specification to produce dynamic output depending on the content of a web page. Natural language generation can automate a significant part of textual production, only requiring a human to supply some important aspects and thus saving considerable time for producing consistent grammatically correct output. In recent years, tools such as SimpleNLG (Gatt and Reiter, 2009) facilitated text realization by a programmer provided they program their application in Java. This system was then extended with SimpleNLG-EnFr (Vaudry and Lapalme, 2013), a English-French version of SimpleNLG. Another approach to text realization is JSreal (Daoust and Lapalme, 2014), a French Web realizer written in JavaScript. This paper describes an attempt at combining the ideas of SimpleNLGEnFr and JSreal to produce a bilingual realizer for French and English from a single specification. JSrealB generates well-formed expressions and sentences. It can be used standalone for linguistic demonstrations or be integrated into complex text generation projects. But like JSreal, it is aimed at web developers, from taking care of morphology, declension and conjugation to creating well-formed texts. A web programmer who wishes to use JSrealB to produce flexible English and/or French textual or HTML output only needs to add two lines in the page: one for importing program and one for calling JSrealB loader to load the resources (i.e. lexicon and rules). The principles underlying JSrealB are similar to those of SimpleNLG: programming language instructions create data structures corresponding to the constituents of the sentence to be produced. Once the data structure (a tree) is built in memory, it is traversed to produce the list of tokens of the sentence. This data structure is built by function calls whose names are the same as the symbols usually used for classical syntax trees: for example, N to create a noun structure, NP for a Noun Phrase, V for a Verb, D for a determiner, S for a Sentence and so on. Features added to the structures using the dot notation can modify the values according to what is intended. JSrealBsyntactic representation is patterned after classical constituent grammar notations. For example, S(NP(D(\"a\"),N(\"woman\")).n(\"p\"), VP(\"eat\").t(\"ps\")) is the JSrealB specification for The women ate. Plural is indicated with feature n(\"p\") where n indicates number and p plural. The verb is conjugated to past tense indicated by the feature tense t and value ps. Agreement between NP and VP is performed automatically. French and English are languages whose structures are similar. Both languages use the same alphabet, they are bot","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124988856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We investigate the task of predicting prepositions that can be used to describe the spatial relationships between pairs of objects depicted in images. We explore the extent to which such spatial prepositions can be predicted from (a) language information, (b) visual information, and (c) combinations of the two. In this paper we describe the dataset of object pairs and prepositions we have created, and report first results for predicting prepositions for object pairs, using a Naive Bayes framework. The features we use include object class labels and geometrical features computed from object bounding boxes. We evaluate the results in terms of accuracy against human-selected prepositions.
{"title":"Generating Descriptions of Spatial Relations between Objects in Images","authors":"A. Muscat, A. Belz","doi":"10.18653/v1/W15-4717","DOIUrl":"https://doi.org/10.18653/v1/W15-4717","url":null,"abstract":"We investigate the task of predicting prepositions that can be used to describe the spatial relationships between pairs of objects depicted in images. We explore the extent to which such spatial prepositions can be predicted from (a) language information, (b) visual information, and (c) combinations of the two. In this paper we describe the dataset of object pairs and prepositions we have created, and report first results for predicting prepositions for object pairs, using a Naive Bayes framework. The features we use include object class labels and geometrical features computed from object bounding boxes. We evaluate the results in terms of accuracy against human-selected prepositions.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115575936","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Typically, human evaluation of NLG output is based on user ratings. We collected ratings and reading time data in a simple, low-cost experimental paradigm for text generation. Participants were presented corpus texts, automatically linearised texts, and texts containing predicted referring expressions and automatic linearisation. We demonstrate that the reading time metrics outperform the ratings in classifying texts according to their quality. Regression analyses showed that self-reported ratings discriminated poorly between the kinds of manipulation, especially between defects in word order and text coherence. In contrast, a combination of objective measures from the low-cost mouse contingent reading paradigm provided very high classification accuracy and thus, greater insight into the actual quality of an automatically generated text.
{"title":"Reading Times Predict the Quality of Generated Text Above and Beyond Human Ratings","authors":"Sina Zarrieß, Sebastian Loth, David Schlangen","doi":"10.18653/v1/w15-4705","DOIUrl":"https://doi.org/10.18653/v1/w15-4705","url":null,"abstract":"Typically, human evaluation of NLG output is based on user ratings. We collected ratings and reading time data in a simple, low-cost experimental paradigm for text generation. Participants were presented corpus texts, automatically linearised texts, and texts containing predicted referring expressions and automatic linearisation. We demonstrate that the reading time metrics outperform the ratings in classifying texts according to their quality. Regression analyses showed that self-reported ratings discriminated poorly between the kinds of manipulation, especially between defects in word order and text coherence. In contrast, a combination of objective measures from the low-cost mouse contingent reading paradigm provided very high classification accuracy and thus, greater insight into the actual quality of an automatically generated text.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125815993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Usage based car insurances, which use sensors to track driver behaviour, are enjoying growing popularity. Although the data collected by these insurances could provide detailed feedback about the driving style, this information is usually kept away from the driver and is used only to calculate insurance premiums. In this paper, we explored the possibility of providing drivers with textual feedback based on telemetric data in order to improve individual driving, but also general road safety. We report that textual feedback generated through NLG was preferred to non-textual summaries currently popular in the field and specifically was better at giving users a concrete idea of how to adapt their driving.
{"title":"Creating Textual Driver Feedback from Telemetric Data","authors":"Daniel Braun, Ehud Reiter, Advaith Siddharthan","doi":"10.18653/v1/W15-4726","DOIUrl":"https://doi.org/10.18653/v1/W15-4726","url":null,"abstract":"Usage based car insurances, which use sensors to track driver behaviour, are enjoying growing popularity. Although the data collected by these insurances could provide detailed feedback about the driving style, this information is usually kept away from the driver and is used only to calculate insurance premiums. In this paper, we explored the possibility of providing drivers with textual feedback based on telemetric data in order to improve individual driving, but also general road safety. We report that textual feedback generated through NLG was preferred to non-textual summaries currently popular in the field and specifically was better at giving users a concrete idea of how to adapt their driving.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115228716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Alignment of interlocutors is a well known psycholinguistic phenomenon of great relevance for dialogue systems in general and natural language generation in particular. In this paper, we present the alignment-capable microplanner SPUD prime. Using a priming-based model of interactive alignment, it is flexible enough to model the alignment behaviour of human speakers to a high degree. This will allow for further investigation of which parameters are important to model alignment and how the human--computer interaction changes when the computer aligns to its users.
{"title":"An Alignment-Capable Microplanner for Natural Language Generation","authors":"Hendrik Buschmeier, K. Bergmann, S. Kopp","doi":"10.3115/1610195.1610207","DOIUrl":"https://doi.org/10.3115/1610195.1610207","url":null,"abstract":"Alignment of interlocutors is a well known psycholinguistic phenomenon of great relevance for dialogue systems in general and natural language generation in particular. In this paper, we present the alignment-capable microplanner SPUD prime. Using a priming-based model of interactive alignment, it is flexible enough to model the alignment behaviour of human speakers to a high degree. This will allow for further investigation of which parameters are important to model alignment and how the human--computer interaction changes when the computer aligns to its users.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126014259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
D. Byron, Alexander Koller, Kristina Striegnitz, Justine Cassell, R. Dale, Johanna D. Moore, J. Oberlander
We describe the first installment of the Challenge on Generating Instructions in Virtual Environments (GIVE), a new shared task for the NLG community. We motivate the design of the challenge, describe how we carried it out, and discuss the results of the system evaluation.
{"title":"Report on the First NLG Challenge on Generating Instructions in Virtual Environments (GIVE)","authors":"D. Byron, Alexander Koller, Kristina Striegnitz, Justine Cassell, R. Dale, Johanna D. Moore, J. Oberlander","doi":"10.3115/1610195.1610223","DOIUrl":"https://doi.org/10.3115/1610195.1610223","url":null,"abstract":"We describe the first installment of the Challenge on Generating Instructions in Virtual Environments (GIVE), a new shared task for the NLG community. We motivate the design of the challenge, describe how we carried it out, and discuss the results of the system evaluation.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129803280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In a corpus study we found that authors vary both mathematical form and precision when expressing numerical quantities. Indeed, within the same document, a quantity is often described vaguely in some places and more accurately in others. Vague descriptions tend to occur early in a document and to be expressed in simpler mathematical forms (e.g., fractions or ratios), whereas more accurate descriptions of the same proportions tend to occur later, often expressed in more complex forms (e.g., decimal percentages). Our results can be used in Natural Language Generation (1) to generate repeat descriptions within the same document, and (2) to generate descriptions of numerical quantities for different audiences according to mathematical ability.
{"title":"Precision and Mathematical Form in First and Subsequent Mentions of Numerical Facts and their Relation to Document Structure","authors":"Sandra Williams, R. Power","doi":"10.3115/1610195.1610215","DOIUrl":"https://doi.org/10.3115/1610195.1610215","url":null,"abstract":"In a corpus study we found that authors vary both mathematical form and precision when expressing numerical quantities. Indeed, within the same document, a quantity is often described vaguely in some places and more accurately in others. Vague descriptions tend to occur early in a document and to be expressed in simpler mathematical forms (e.g., fractions or ratios), whereas more accurate descriptions of the same proportions tend to occur later, often expressed in more complex forms (e.g., decimal percentages). Our results can be used in Natural Language Generation (1) to generate repeat descriptions within the same document, and (2) to generate descriptions of numerical quantities for different audiences according to mathematical ability.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128884174","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The GREC Task at REG '08 required participating systems to select coreference chains to the main subject of short encyclopaedic texts collected from Wikipedia. Three teams submitted a total of 6 systems, and we additionally created four baseline systems. Systems were tested automatically using a range of existing intrinsic metrics. We also evaluated systems extrinsically by applying coreference resolution tools to the outputs and measuring the success of the tools. In addition, systems were tested in a reading/comprehension experiment involving human subjects. This report describes the GREC Task and the evaluation methods, gives brief descriptions of the participating systems, and presents the evaluation results.
{"title":"The TUNA-REG Challenge 2009: Overview and Evaluation Results","authors":"Albert Gatt, A. Belz, Eric Kow","doi":"10.3115/1610195.1610224","DOIUrl":"https://doi.org/10.3115/1610195.1610224","url":null,"abstract":"The GREC Task at REG '08 required participating systems to select coreference chains to the main subject of short encyclopaedic texts collected from Wikipedia. Three teams submitted a total of 6 systems, and we additionally created four baseline systems. Systems were tested automatically using a range of existing intrinsic metrics. We also evaluated systems extrinsically by applying coreference resolution tools to the outputs and measuring the success of the tools. In addition, systems were tested in a reading/comprehension experiment involving human subjects. This report describes the GREC Task and the evaluation methods, gives brief descriptions of the participating systems, and presents the evaluation results.","PeriodicalId":307841,"journal":{"name":"European Workshop on Natural Language Generation","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130966848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：