Showing papers on "Chatbot published in 2005"

Using corpora in machine-learning chatbot systems

Abstract: A chatbot is a machine conversation system which interacts with human users via natural conversational language. Software to machine-learn conversational patterns from a transcribed dialogue corpus has been used to generate a range of chatbots speaking various languages and sublanguages including varieties of English, as well as French, Arabic and Afrikaans. This paper presents a program to learn from spoken transcripts of the Dialogue Diversity Corpus of English, the Minnesota French Corpus, the Corpus of Spoken Afrikaans, the Qur’an Arabic-English parallel corpus, and the British National Corpus of English; we discuss the problems which arose during learning and testing. Two main goals were achieved from the automation process. One was the ability to generate different versions of the chatbot in different languages, bringing chatbot technology to languages with few if any NLP resources: the corpus-based learning techniques transferred straightforwardly to develop chatbots for Afrikaans and Qur’anic Arabic. The second achievement was the ability to learn a very large number of categories within a short time, saving effort and errors in doing such work manually: we generated more than one million AIML categories or conversation-rules from the BNC corpus, 20 times the size of existing AIML rule-sets, and probably the biggest AI Knowledge-Base ever.

Freudbot: An Investigation of Chatbot Technology in Distance Education

Abstract: A chatbot named Freudbot was constructed using the open source architecture of AIML to determine if a famous person application of chatbot technology could improve student-content interaction in distance education. Fifty-three students in psychology completed a study in which they chatted with Freudbot over the web for 10 minutes under one of two instructional sets. They then completed a questionnaire to provide information about their experience and demographic variables. The results from the questionnaire indicated a neutral evaluation of the chat experience although participants positively endorsed the expansion of chatbot technology and provided clear direction for future development and improvement. A basic analysis of the chatlogs indicated a high proportion of on-task behaviour. There was no effect of instructional set. Altogether, the findings indicate that famous person applications of chatbot technology may be promising as a teaching and learning tool in distance and online education. Chatbots are agents programmed to mimic human conversationalists. The first and still quite successful chatbot was ELIZA (Weizenbaum, 1966), a computer program designed to emulate a Rogerian therapist, a type of self-directed therapy where the patient’s discourse is redirected back to the patient by the therapist usually in the form of a question. “Its name was chosen to emphasize that it may be incrementally improved by its users, since its language abilities may be continually improved by a "teacher". Like the ELIZA of Pygmalion fame, it can be made to appear even more civilized, the relation of appearance to reality, however, remaining in the domain of the playwright.” (Weizenbaum, 1966, p.2) The playwright in this case is the programmer but instead of classic Artificial Intelligence, ELIZA was programmed with rules to give the illusion of understanding. Essentially, ELIZA was programmed to recognize keywords and choose an appropriate transformation based on the immediate linguist context. Weizenbaum used the term ‘script’ to refer to the collection of keywords and associated transformation rules. Even though ELIZA is easily exposed as a fraud in the Turing sense, the popularity of the Rogerian therapist script remains high and there are a number of sites that allow you access to ELIZA. It is interesting to note that of all the scripts planned and developed by Weisenbaum, the Rogerian therapist script was the most enduring. Arguably the most successful chatbot today is ALICE (Artificial Linguistic Internet Chat Entity), 3 time winner of the Loebner Prize, the holy grail for chatbots. ALICE was written by Richard Wallace and although no chatbot has passed the Turing test in the Loebner competition, ALICE has been judged the most human-like chatbot in 2000, 2001, and 2004. Like ELIZA, ALICE has no true understanding and is programmed to recognize templates and respond with patterns according to the context. Moreover, like ELIZA, ALICE is incrementally improved with the addition of new responses. Unlike ELIZA, ALICE is programmed to talk to people on the web for as long as possible on any topic. Compared to the ELIZA’s knowledge of 200 keywords and rules, ALICE is embodied by approximately 41,000 templates and associated patterns. Perhaps the most important difference between ALICE and ELIZA is that ALICE is written in AIML (Artificial Intelligence Markup Language), an XML-based open source language with a reasonably active development community. There are also a variety of AIML parsers available written in Java, Perl, PHP, and C++ that permit interaction through a variety of interfaces, from simple web pages to Flash-based (or other) animation, instant messaging, and even voice input/output. In addition, Pandorabots, a web service that promotes and supports the use of ALICE and AIML is reporting support for over 20,000 chatbots on their site (http://www.pandorabots.com). At Pandorabots, would-be botmasters can easily create their own chatbot by modifying the personality of ALICE or by starting from scratch. An AIML chatbot is suitable for many educational applications but our interest was in the famous personality application. Specifically, we were interested in whether students would enjoy and benefit from chatting with famous historical figures in psychology. As a distance education provider, we are always looking for ways to improve the interaction between student and course content over the web. Chatting with an historical figure via the internet may be intrinsically more interesting than the same information presented in a standard third party format over the web. In terms of a theoretical rationale, there are several bases for investigating a famous personality application of chatbot technology as learning tool in distance education. Social constructionist theories of learning emphasize collaboration and conversation as a natural and effective means of knowledge construction and elaboration. The work of Graesser and colleagues on AutoTutor is based largely on these theories (see Graesser,Wiemer-Hastings, Wiemer-Hastings, Kreuz, & Tutoring Research Group 1999). A second rationale is found in the work of Cassell and colleagues on Embodied Conversational Agents (ECA). Cassell indicates that motivation for their research is based on the primacy of conversation as a natural skill learned early and effortlessly in life (Cassell, Bickmore, Campbell, Vilhjalmsson, & Yan, 2000). A conversational interface to a famous psychologist should be engaging and intuitive. A third rationale is provided through cognitive resource theory that argues linguistic rules governing conversational exchanges are automatic in nature due to frequency of use and consequently, free up additional resources to devote to encoding, understanding, and learning. Finally, according to the media equation (Reeves & Nass, 1996), people are predisposed to treat computers, television and other instances of media as people. They describe a number of experimental studies that generally show no differences in how media is ‘treated’ in comparison to people. The social rules that govern human-human interactions appear to govern human-computer interactions as well. If this is the case, then people may be predisposed to interact with a famous person application on the computer given the close fit of the application to human and conversational characteristics.

FAQchat as in Information Retrieval system

Abstract: A chatbot is a conversational agent that interacts with users through natural languages. In this paper, we describe a new way to access information using a chatbot. The FAQ in the School of Computing at the University of Leeds has been used to retrain the ALICE chatbot system, producing FAQchat. The results returned from FAQchat are similar to ones generated by search engines such as Google. For evaluation, a comparison was made between FAQchat and Google. The main objective is to demonstrate that FAQchat is a viable alternative to Google and it can be used as a tool to access FAQ databases.

A chatbot system as a tool to animate a corpus

Abstract: eprints@whiterose.ac.uk https://eprints.whiterose.ac.uk/ Reuse Unless indicated otherwise, fulltext items are protected by copyright with all rights reserved. The copyright exception in section 29 of the Copyright, Designs and Patents Act 1988 allows the making of a single copy solely for the purpose of non-commercial research or private study within the limits of fair dealing. The publisher or other rights-holder may allow further reproduction and re-use of this version refer to the White Rose Research Online record for this item. Where records identify the publisher as the copyright holder, users can verify any specific terms of use on the publisher’s website.

Computing Machinery and the Individual: the Personal Turing Test

Abstract: What are the key features of verbal and non-verbal communication that make a person not just any person, but that person? We pose this challenge in the context of an advancement to the Turing Test. If we know the answers to this question, the key features for any specific individual could be embodied within computerised representations, or agents. For a computerised agent convincingly to represent a real person to that person’s contacts, friends, family and colleagues is likely to have many applications as we move to the era of the self distributed over networks. Here we lay out the key features of our new Personal Turing Test, with reference to a highly successful – and convincing – (verbal) chatbot, called Jabberwacky. We expand our ideas somewhat to cover non verbal cues that may characterise a person, and end with suggested application examples and ethical questions.

A corpus based approach to generalise a chatbot system

Abstract: Chatbot tools are computer programs which interact with users using natural languages. Most developers built their systems aiming to fool users that they are talking with real humans. Up to now most chatbots serve as a tool to amuse users through chatting with a robot. However, the knowledge bases of almost all chatbots are edited manually which restricts users to specific languages and domains. This thesis shows that chatbot technology could be used in many different ways in addition to being a tool for fun. A chatbot could be used as a tool to learn or study a new language, a tool to access an information system, a tool to visualise the contents of a corpus and a tool to give answers to questions in a specific domain. Instead of being restricted to a specific domain or written language, a chatbot could be trained with any text in any language. Some of the differences between real human conversations and human-chatbot dialogues are presented. A Java program has been developed to read a text from a machine readaatbble text (corpus) and convert it to ALICE chatbot format language (AIML). The program was built to be general, the generality in this respect implies no restrictions on specific language, domain or structure. Different languages were tested: English, Arabic, Afrikaans, French and Spanish. At the same time different corpora structures were used: dialogue, monologue and structured text.