In order to facilitate efficient creation of instructional content for high proficiency language learning systems, we investigate the use of utterance templates through an intuitive authoring tool. Evaluation in the context of authoring mini-dialogs shows that the tool can help authors can achieve higher coverage of the target language.

Human tutors, in dealing with non-native speakers, draw from their knowledge of common learner mistakes to understand learner speech and offer effective corrective advice. In this paper we present our work towards embedding some of this knowledge in the speech recognition and learner speech error detection subsystems of the Tactical Language Training System (TLTS). We discuss the implementation and effectiveness of our methodology which uses a combination of rule based and probabilistic models derived from linguistic knowledge about the target language and annotated speech to identify potential learner errors, detect them using ASR and to provide the user with corrective feedback based on error severity and factors such as the learner history. Evaluation results show that our system can provide effective feedback to the learner with good accuracy.

Recent research shows that instructors commonly use politeness strategies to achieve affective scaffolding in educational contexts. The importance of affective factors such as self-confidence and interest that contribute to learner motivation is well recognized. In this paper, we describe the results of a Wizard-of-Oz experiment to study the effect of politeness strategies on both cognitive and motivational factors. We compare the results of two different politeness strategies, direct and polite, in assisting seventeen students in a computer-based learning task. We find that politeness can affect students' motivational state and help students learn difficult concepts. The results of the experiment provide a basis for the design of a polite pedagogical agent and its tutorial intervention strategies.

In this paper, we describe a method for pedagogical agents to choose when to interact with learners in interactive learning environments. This method is based on observations of human tutors coaching students in on-line learning tasks. It takes into account the focus of attention of the learner, the learner's current task, and expected time required to perform the task. A Bayesian network model combines evidence from eye gaze and interface actions to infer learner focus of attention. The attention model is combined with a plan recognizer to detect different types of learner difficulties such as confusion and indecision which warrant intervention. We plan to incorporate this capability into a pedagogical agent able to interact with learners in socially appropriate ways.

Guidebots, or animated pedagogical agents, can enhance interactive learning environments by promoting deeper learning and improve the learner's subjective experience. Guidebots exploit a person's natural tendency to interact socially with computers, as documented by Reeves, Nass, and their colleagues. However they also raise expectations of social abilities, and failure to meet those expectations can have unintended negative effects. The Social Intelligence Project is developing improved social interaction skills for guidebots. This paper describes efforts to model and implement interaction tactics for guidebots, i.e., dialog exchanges that are intended to achieve particular communicative and motivational effects. These are based on analyses of student-tutor interaction during computer-based learning

Animated pedagogical agents offer promise as a means of making computer-aided learning more engaging and effective. To achieve this, an agent must be able to interact with the learner in a manner that appears believable, and that furthers the pedagogical goals of the learning environment. In this paper we describe how the user interaction model of one pedagogical agent evolved through an iterative process of design and user testing. The pedagogical agent Adele assists students as they assess and diagnose medical and dental patients in clinical settings. We describe the results of, and our responses to, three studies of Adele, involving over two hundred and fifty medical and dental students over five years, that have led to an improved tutoring strategy, and discuss the interaction possibilities of two different reasoning engines. With the benefit of hindsight, the paper articulates the principles that govern effective user-agent interaction in educational contexts, and describes how the agents interaction design in its current form embodies those principles

Animated pedagogical agents [ire mimated characters that c~ii iriterrict with learners iri corripiiter-based erivirorinietits, iri order to stirriiikite arid encoi4rcige lerirriitig. They improve the &“ectiveriess of education arid trriiriirig applicatioris, arid caii ulso be eniployed iri riiarij’ other iriteractive applicatioris to assist iiovice irsers. They draw 011 techriiqrres froin interactive eritertcririnierit, tnrikirig tlierii rrdl miter1 ,for entertaintnent rippliccitions brith uti ediicatiorirrl pirrpose ( “ediituirinient“). This tritoriril provides NII introdirctiori to anininted pedagogical agents, their itse in education, training, arid eritertainnient applications, arid methods arid technologies used to develop them. The iniportant capabilities of APAS will be described arid illirstrated in the context of exiniple pedagogical agent systems. Coriirnori niethods arid architectures for bidding arid coritrollirig the bekcivior of sirch agents will be described arid contrasted. Empirical evalim tioris of effectiveriess of APAS will also be presented. Particirlar eniphasis ill be placed on cissessirig the ciirrerit state of the art in APA research, highlightirig recerit accoriiplishnierits and iderit(fiirig cireas of ciirrerit Litid fictitre research.

No abstract available.

The main factors that affect software understanding are the complexity of the problem solved by the program, the program text, the user's mental ability and experience and the task being performed. The paper describes a planning approach solution to the software understanding problem that focuses on the user's task and expertise. First, user questions about software artifacts have been studied and the most commonly asked questions are identified. These questions are organized into a question model and procedures for answering them are developed. Then, the patterns in user questions while performing certain tasks have been studied and these patterns are used to build generic task models. The explanation system uses these task models in several ways. The task model, along with a user model, is used to generate explanations tailored to the user's task and expertise. In addition, the task model allows the system to provide explicit task support in its interface.

The approach to representation and presentation of knowledge used in ARIES, an environment to experiment with support for analysts in modeling target domains and in entering and formalizing system requirements, is described. To effectively do this, ARIES must manage a variety of notations so that analysts can enter information in a natural manner, and ARIES can present it back in different notations and from different viewpoints. To provide this functionality, a single, highly expressive internal representation is used for all information in the system. The system architecture separates representation and presentation, in order to localize consistency and propagation issues. The presentation architecture is tailored to be flexible enough so that new notations can be easily introduced on top of the underlying representation. Presentation knowledge is coupled to specification evolution knowledge thereby leveraging common representations for both in order to provide automated focusing support to users who need informative guidance in creating and modifying specifications.>

This paper describes a program called PROUST which does on-line analysis and understanding of Pascal written by novice programmers. PROUST takes as input a program and a nonalgorithmic description of the program requirements, and finds the most likely mapping between the requirements and the code. This mapping is in essence a reconstruction of the design and implementation steps that the programmer went through in writing the program. A knowledge base of programming plans and strategies, together with common bugs associated with them, is used in constructing this mapping. Bugs are discovered in the process of relating plans to the code; PROUST can therefore give deep explanations of program bugs by relating the buggy code to its underlying intentions.