字幕表 動画を再生する 英語字幕をプリント One of the main uses of the principles of natural language processing is what is called “chat bot” or “chatter bot”. Actually the term “chatterBot” was originally coined by Michael Mauldin, the creator of this Verbot, Julia, in 1994 to describe these conversational programs. A chatter bot is a computer program designed to simulate a conversation with human users. The primary aim is to kind of fool the user into thinking that he actually talks with a human. This is what we have already referred to “Turing test”. This new criterion of intelligence soon became a challenge to several specialists, that created all kind of chat bots until today. The first chat bot was Eliza and she was built by Joseph Weizebaum. She was released to the public in 1966 and she played the role of a psychotherapist. Eliza would simply rearrange a submitted statement or question, into a new question to ask the person talking to her. ELIZA's key method of operation, which is now used by all chat bots, involves the recognition of key words or phrases in the input that are previously categorized to positive or negative structures. That way she is creating an illusion of perfect understanding in most cases. At her time, people would actually use Eliza to confide in. This is a conversation we made with Eliza, and as you can see, you can ask her something and answers you back, most of the times with a question. If you repeat yourself, she will probably answer you with the same way. So it is clear that she matches words in order to answer. Parry was next, designed by Kenneth Colby in 1971. His purpose was to reflect the mind of a seriously paranoid mental patient. To achieve that, a database would be created with positive or negative values to certain words or phrases to act as triggers for various emotions. Both Parry and Eliza were successful in playing their roles since professional psychologists were having difficulty to find out that they were actually talking with a software. These chatbots are not self learning A.I. in any way, because their creators spent a lot of time, constructing a fully edited database with words and meanings. This method is used until now, but with some differences. Now, they are also using the bots’ input and output in order to analyze that and find out what is commonly used and what isn’t. So moving on in our timeline, we find Jabberwacky, a kind of different chat bot which claims that it can learn. It was created in 1988 by Rollo Carpentor and brought online in 1997. The “general AI” of Jabberwacky stores everything everyone has ever said, and finds the most appropriate thing to say, using contextual pattern matching techniques. In its website claims that it can be taught slang English, word games, jokes etc, and that everyone can contributes to its learning. They are comparing Jabberwacky with a conversational Wikipedia. And as you can see, we have made a conversation with Jabberwacky as well. One of the most famous chat bots is A.L.I.C.E.. She was created by Dr. Richard Wallance in 1995 using his own A.I.M.L. markup language. ALICE has the structure as Eliza and Parry when it comes to creation. They use human edited database for their responses. But Alice was not created for a purpose, but to demonstrate a woman that you can make a conversation with. This is a huge project since ALICE would have to response to all kind of questions and subjects. But she is considered a strong competitor. There is a long list of chatter bots that can be used in different ways and industries. These days, a lot of companies all over the world, choose to use chat bots in order to guide their websites’ users, or answer their clients’ questions. There are chat bots that will answer questions about sports, animals and electronics. Others can give you ideas about fashion or cooking. There are Chat bots that have been used by several governments and financial purposes. And yes, there are erotic chat bots that can answer to all sort of questions, of course you have to be an adult! All of these chat bots work for several organizations in order to make things easier with information and to guide clients or users through services. They exist through the internet and maybe you have someday used one, and haven’t noticed. Chat bots are often divided by their appearance, perception and expressions. So there are faceless chat bots, robots, avatars and animated avatars and several search boxes. Some can recognize text, others can recognize speech and respond with speech as these examples, and others can recognize gestures and make facial expressions. For example this Polish chat bot, where we asked her if she knows English and she nodded her head, saying no. Some chat bots can be used for one application only and they are created for one purpose. Maybe e-learning, maybe gaming, navigation through a website, answering questions for a shop’s products or just entertainment. There are actually some amazing chat bots that offer a lot of different features to their users and a lot of help through their everyday life. One of these complex and exciting application is Siri (which stands for Speech Interpretation and Recognition Interface) which was founded in 2007 by Dag Kittlaus, Adam Cheyer, Tom Gruber and Norman Winarsky. Siri is an intelligent personal assistant and also a knowledge navigator which works as an application for a certain mobile operating system. Siri uses natural language user interface, to understand the commands of the user. Combining other application that the mobile offers and the amazing understanding of natural language, she is able to answer questions, make recommendations and perform actions by delegating requests to a set of web services. Now, let’s have some fun with the famous character Raj, from “The Bing Bang Theory” series, using Siri in a humorous way. Despite the greatness of chat bots and this kind of applications, the expanding of technology, inspires some minds in order to use it maliciously. Those chat bots are frequently used to fill chat rooms with spam and advertising, or to make people reveal personal information such as bank account numbers and passwords. So there must be a great caution when using one and be careful if something is suspicious when we talk to someone by instant messaging. So, maybe chat bots seem to be just luxury since, they perform commands that we could perform on our own and also they may not respond always as appropriate as a person would do. We must not forget though, that their main purpose of creation is to serve clients and users as well as inform them and help them throughout their everyday tasks. So, maybe chat bots are not fully included in our daily life yet, but who knows? Maybe one day we would not be able to resist the help of a chat bot or should we say a new friend?
A2 初級 自然言語処理とチャットボット PART B (Natural Language Processing and Chat Bots PART B) 58 3 shin に公開 2021 年 01 月 14 日 シェア シェア 保存 報告 動画の中の単語