Adaptive User Interfaces for Information Extraction

Total Page:16

File Type:pdf, Size:1020Kb

Adaptive User Interfaces for Information Extraction From: AAAI Technical Report SS-00-01. Compilation copyright © 2000, AAAI (www.aaai.org). All rights reserved. Adaptive User Interfaces for Information Extraction Peter Vanderheyden and Robin Cohen Department of Computer Science, University of Waterloo Waterloo, Ontario, Canada Abstract Information Extraction: What Is It? Informationextraction systemsdistill specific items of Information extraction involves "extraction or pulling informationfrom large quantities of text. Current lit- out of pertinent information from large volumes of erature in the field of information extraction focuses texts" (MUC-71999). It is often defined by example, largely on howdifferent language models and machine terms of the results the system would return for a given learning approaches can improve system accuracy, with query, and often contrasted with what an information little mention of the role of interaction betweenthe user and the system. In this paper, we explore many retrieval system might return for a similar query. opportunities for interaction and adaptation during in- formation extraction, and propose a system design to The query-result perspective investigate these opportunities further. Information retrieval generally interprets a query as a string of unrelated words or word phrases 2, matching Introduction it against the words in the corpus documents. Output Muchhas been written about the flood of information from the IR system consists of either whole documents currently available, and the challenge of locating rel- or segments, deemedrelevant to the query words. Thus, evant items amid that flood. Several areas of applied for P’s question "Is terrorism on the rise?", if the corpus research -- information extraction (IE), information re- contains a document that mentions a ’rise’ in ’terror- trieval (IR), and knowledge discovery in text (KDT) ism’, and assuming P has confidence in the document, seek to address this challenge. Weare specifically in- such a document will have answered P’s question. terested in the field of information extraction. In this An IE system takes as its input a user’s query ex- paper, we discuss opportunities for interaction between pressed (or expressable) as a sentence template, con- user and system, and for adaptation by the system to taining unfilled slots, and relations betweenthose slots. facilitate this interaction. Wethen develop objectives For example, a query to identify terrorism events con- for designing a system that supports the user effectively tained in a document could be written as: during information extraction. "A terrorism event occurred in which {GroupA} Wemotivate this discussion by observing that little {ActB} {GroupB}, and {Acre} {0bjectC}." is mentioned in current literature about the interaction between an information extraction system and a user. The system then processes each document in a corpus In contrast, for example, information retrieval litera- and, for each occurrence in the document of an event ture discusses interactions between users and computer- that adequately fills the query template, the system re- based systems as well as between human librarians and turns the filled template (e.g., (ActB}="kidnapped’, library patrons (Twidale & Nichols 1998). {GroupS}="10 civilians", etc.). Given P’s question, The following will be an ongoing example to which converted into an appropriate template form, the sys- we refer throughout the paper: Person P is concerned tem would ideally return all terrorism events mentioned that there has been a high numberof terrorism cases re- in the corpus. Thus one possible scenario is that, if the ported recently t, and is looking for more information on IR system failed to answer P’s question satisfactorily, the matter. What P really wants to know (P’s "global the IE system could be used as a follow-up approach information need") is the following: "Is terrorism on (Gaizauskas ~ Robertson 1997). The data returned the rise?" System S is able to accept many different by the IE system could be analyzed (the number of kinds of queries, and then to search for information in events counted) and plotted as a function of time, to a corpus of documents. How should P and S interact to satisfy P’s information need? 2Most current information retrieval systems consider only minor variations (e.g., morphological) on the query XThereader mayinsert any other topic in place of "ter- words, though natural language approaches to IR exist as rorism" (e.g., managementsuccessions, cancer, rap music). well (Lewis & Sparck Jones 1996). 136 see whether terrorism cases were indeed increasing in user to another, or within the span of a single task; this frequency. is by no meansan exhaustive list. It is not our intention Knowledge discovery in text (Feldman & Dagan to suggest that any one system should explicitly make 1995) begins with the kind of well-structured data often use of all of these parameters. However, the designers returned by an information extraction system. The sys- of a system should at least be aware of them, in order tem and user then cooperate, the system using statisti- to take into account any that are relevant to the specific cal calculations and the user using intuition and knowl- design and intended application of a given system. edge of the world, to find correlative patterns among Many parameters mentioned in this section come data categories, possibly defining new categories in the from (Allen 1996) and are not specific to information ex- process. For example, a KDTsystem might analyze traction, but are relevant to various information tasks. the terrorist event data returned by the IE query men- tioned earlier, to identify patterns of occurrence over Information-based variations the a range of years. Thus S would compute the an- What kind of information is the user interested in? swer to P’s question on its own, rather than relying What kind of information is contained in the docu- on a document (that P trusts) in the corpus contain- ments? Different kinds of information may not be con- ing the answer. Finding information often does involve ducive to the same techniques for representation or in- trying one approach, getting partial answers, integrat- teraction. (For example, diagrammatic information is ing this new knowledge, and trying again with the same more successfully conveyed in a diagram than as text.) approach or with a different one. Whydoes the user want the information that the system is trying to find? For example, is the user: The natural language perspective ¯ searching for something in particular; Some degree of language understanding is required to recognize how the terms in the query template are re- ¯ generally browsing, or exploring, to become ac- lated to one another. The documents in the corpus quainted with information available on the topic; maycontain either unstructured natural language (e.g., ¯ integrating new information into what the user al- news wire articles, in the MUCs;e.g., (MUC-61995) ready knows; or, and (MUC-71999)) or semi-structured text (e.g., ¯ trying to gain new perspectives on what the user al- mulaic reviews of restaurants or movies, with consistent ready know? structure; discussed in manyof the papers in (AAAI’98- AIII 1998)). When documents contain unstructured The user’s motivation for performing the IE task will natural language, information extraction can be seen as affect how she examines each document and what kind a natural language understanding task restricted to the of information she focuses on. As a result, it mayaffect domain of knowledge relevant to the user’s query. The howthe user prefers to interact with the system. system typically contains a knowledge base of words Howmany solutions to the query is the user looking and phrases, semantic relations, and rules of syntax in for? If the user is interested in finding a small number the language. Howthese syntax rules are expressed is of solutions then a less accurate (and possibly faster) very muchaffected by the specific language model being procedure may suffice. This is in contrast to a user used (e.g., finite state automata (Appelt et ai. 1995); interested in all solutions that occur in the corpus, or statistical dependency model (Weischedel 1995)). for the "best" solution (which may require finding all Information extraction systems acquire additional solutions, then using somecriterion to select the best). knowledge (about the lexicon, semantics, etc.) dur- Similarly, is the user satisfied with finding someof the ing the course of the task through machine learning elements in the query (i.e., in information extraction, techniques. Typically, this involves "batch" learning -- this means that partially filled templates are accept- learning in large increments -- with no user involvement able)? If so, how many and which elements (if they are within each batch, and the specific approach taken to not weighted equally in importance) are required? will be very much dependent on the language model. With respect to documents in the corpus, does it mat- ter how documents are related to one another? For ex- Summary: there’s still something missing ample, in the legal domain it is important to relate a given case to possible precedents (Rose & Belew 1991). A good deal of the process involved in information ex- traction has yet to be discussed. For example, the de- User-based variations sign of an information extraction system should con- Howdoes the user process different kinds of informa- sider the kinds of users, queries, and corpora that will be involved. And information extraction may occur in tion? In any given situation, different users will vary with respect to: isolation, but in the context of other activities. ¯ what the user perceives, identifies, or focuses on3; The Information Task In Context 3Allen (1996) discusses individuals’ "boundedrational- In this section, we briefly mention some of the param- ity", the fact that people do not always make use of all eters that will vary from one task to another, from one available information-- i.e., real people are not ideal users.
Recommended publications
  • Information Extraction from Electronic Medical Records Using Multitask Recurrent Neural Network with Contextual Word Embedding
    applied sciences Article Information Extraction from Electronic Medical Records Using Multitask Recurrent Neural Network with Contextual Word Embedding Jianliang Yang 1 , Yuenan Liu 1, Minghui Qian 1,*, Chenghua Guan 2 and Xiangfei Yuan 2 1 School of Information Resource Management, Renmin University of China, 59 Zhongguancun Avenue, Beijing 100872, China 2 School of Economics and Resource Management, Beijing Normal University, 19 Xinjiekou Outer Street, Beijing 100875, China * Correspondence: [email protected]; Tel.: +86-139-1031-3638 Received: 13 August 2019; Accepted: 26 August 2019; Published: 4 September 2019 Abstract: Clinical named entity recognition is an essential task for humans to analyze large-scale electronic medical records efficiently. Traditional rule-based solutions need considerable human effort to build rules and dictionaries; machine learning-based solutions need laborious feature engineering. For the moment, deep learning solutions like Long Short-term Memory with Conditional Random Field (LSTM–CRF) achieved considerable performance in many datasets. In this paper, we developed a multitask attention-based bidirectional LSTM–CRF (Att-biLSTM–CRF) model with pretrained Embeddings from Language Models (ELMo) in order to achieve better performance. In the multitask system, an additional task named entity discovery was designed to enhance the model’s perception of unknown entities. Experiments were conducted on the 2010 Informatics for Integrating Biology & the Bedside/Veterans Affairs (I2B2/VA) dataset. Experimental results show that our model outperforms the state-of-the-art solution both on the single model and ensemble model. Our work proposes an approach to improve the recall in the clinical named entity recognition task based on the multitask mechanism.
    [Show full text]
  • The History and Recent Advances of Natural Language Interfaces for Databases Querying
    E3S Web of Conferences 229, 01039 (2021) https://doi.org/10.1051/e3sconf/202122901039 ICCSRE’2020 The history and recent advances of Natural Language Interfaces for Databases Querying Khadija Majhadi1*, and Mustapha Machkour1 1Team of Engineering of Information Systems, Faculty of Sciences Agadir, Morocco Abstract. Databases have been always the most important topic in the study of information systems, and an indispensable tool in all information management systems. However, the extraction of information stored in these databases is generally carried out using queries expressed in a computer language, such as SQL (Structured Query Language). This generally has the effect of limiting the number of potential users, in particular non-expert database users who must know the database structure to write such requests. One solution to this problem is to use Natural Language Interface (NLI), to communicate with the database, which is the easiest way to get information. So, the appearance of Natural Language Interfaces for Databases (NLIDB) is becoming a real need and an ambitious goal to translate the user’s query given in Natural Language (NL) into the corresponding one in Database Query Language (DBQL). This article provides an overview of the state of the art of Natural Language Interfaces as well as their architecture. Also, it summarizes the main recent advances on the task of Natural Language Interfaces for databases. 1 Introduction of recent Natural Language Interfaces for databases by highlighting the architectures they used before For several years, databases have become the preferred concluding and giving some perspectives on our future medium for data storage in all information management work.
    [Show full text]
  • A Meaningful Information Extraction System for Interactive Analysis of Documents Julien Maitre, Michel Ménard, Guillaume Chiron, Alain Bouju, Nicolas Sidère
    A meaningful information extraction system for interactive analysis of documents Julien Maitre, Michel Ménard, Guillaume Chiron, Alain Bouju, Nicolas Sidère To cite this version: Julien Maitre, Michel Ménard, Guillaume Chiron, Alain Bouju, Nicolas Sidère. A meaningful infor- mation extraction system for interactive analysis of documents. International Conference on Docu- ment Analysis and Recognition (ICDAR 2019), Sep 2019, Sydney, Australia. pp.92-99, 10.1109/IC- DAR.2019.00024. hal-02552437 HAL Id: hal-02552437 https://hal.archives-ouvertes.fr/hal-02552437 Submitted on 23 Apr 2020 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. A meaningful information extraction system for interactive analysis of documents Julien Maitre, Michel Menard,´ Guillaume Chiron, Alain Bouju, Nicolas Sidere` L3i, Faculty of Science and Technology La Rochelle University La Rochelle, France fjulien.maitre, michel.menard, guillaume.chiron, alain.bouju, [email protected] Abstract—This paper is related to a project aiming at discover- edge and enrichment of the information. A3: Visualization by ing weak signals from different streams of information, possibly putting information into perspective by creating representa- sent by whistleblowers. The study presented in this paper tackles tions and dynamic dashboards.
    [Show full text]
  • Span Model for Open Information Extraction on Accurate Corpus
    Span Model for Open Information Extraction on Accurate Corpus Junlang Zhan1,2,3, Hai Zhao1,2,3,∗ 1Department of Computer Science and Engineering, Shanghai Jiao Tong University 2 Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering, Shanghai Jiao Tong University, Shanghai, China 3 MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University [email protected], [email protected] Abstract Sentence Repeat customers can purchase luxury items at reduced prices. Open Information Extraction (Open IE) is a challenging task Extraction (A0: repeat customers; can purchase; especially due to its brittle data basis. Most of Open IE sys- A1: luxury items; A2: at reduced prices) tems have to be trained on automatically built corpus and evaluated on inaccurate test set. In this work, we first alle- viate this difficulty from both sides of training and test sets. Table 1: An example of Open IE in a sentence. The extrac- For the former, we propose an improved model design to tion consists of a predicate (in bold) and a list of arguments, more sufficiently exploit training dataset. For the latter, we separated by semicolons. present our accurately re-annotated benchmark test set (Re- OIE2016) according to a series of linguistic observation and analysis. Then, we introduce a span model instead of previ- ous adopted sequence labeling formulization for n-ary Open long short-term memory (BiLSTM) labeler and BIO tagging IE. Our newly introduced model achieves new state-of-the-art scheme which built the first supervised learning model for performance on both benchmark evaluation datasets.
    [Show full text]
  • Information Extraction Overview
    INFORMATION EXTRACTION OVERVIEW Mary Ellen Okurowski Department of Defense. 9800 Savage Road, Fort Meade, Md. 20755 meokuro@ afterlife.ncsc.mil 1. DEFINITION OF INFORMATION outputs informafiou in Japanese. EXTRACTION Under ARPA sponsorship, research and development on information extraction systems has been oriented toward The information explosion of the last decade has placed evaluation of systems engaged in a series of specific increasing demands on processing and analyzing large application tasks [7][8]. The task has been template filling volumes of on-line data. In response, the Advanced for populating a database. For each task, a domain of Research Projects Agency (ARPA) has been supporting interest (i.e.. topic, for example joint ventures or research to develop a new technology called information microelectronics chip fabrication) is selected. Then this extraction. Information extraction is a type of document domain scope is narrowed by delineating the particular processing which capttnes and outputs factual information types of factual information of interest, specifically, the contained within a document. Similar to an information generic type of information to be extracted and the form of retrieval (lid system, an information extraction system the output of that information. This information need responds to a user's information need. Whereas an IR definition is called the template. The design of a template, system identifies a subset of documents in a large text which corresponds to the design of the database to be database or in a library scenario a subset of resources in a populated, resembles any form that you may fdl out. For library, an information extraction system identifies a subset example, certain fields in the template require normalizing of information within a document.
    [Show full text]
  • Open Information Extraction from The
    Open Information Extraction from the Web Michele Banko, Michael J Cafarella, Stephen Soderland, Matt Broadhead and Oren Etzioni Turing Center Department of Computer Science and Engineering University of Washington Box 352350 Seattle, WA 98195, USA {banko,mjc,soderlan,hastur,etzioni}@cs.washington.edu Abstract Information Extraction (IE) has traditionally relied on ex- tensive human involvement in the form of hand-crafted ex- Traditionally, Information Extraction (IE) has fo- traction rules or hand-tagged training examples. Moreover, cused on satisfying precise, narrow, pre-specified the user is required to explicitly pre-specify each relation of requests from small homogeneous corpora (e.g., interest. While IE has become increasingly automated over extract the location and time of seminars from a time, enumerating all potential relations of interest for ex- set of announcements). Shifting to a new domain traction by an IE system is highly problematic for corpora as requires the user to name the target relations and large and varied as the Web. To make it possible for users to to manually create new extraction rules or hand-tag issue diverse queries over heterogeneous corpora, IE systems new training examples. This manual labor scales must move away from architectures that require relations to linearly with the number of target relations. be specified prior to query time in favor of those that aim to This paper introduces Open IE (OIE), a new ex- discover all possible relations in the text. traction paradigm where the system makes a single In the past, IE has been used on small, homogeneous cor- data-driven pass over its corpus and extracts a large pora such as newswire stories or seminar announcements.
    [Show full text]
  • Information Extraction Using Natural Language Processing
    INFORMATION EXTRACTION USING NATURAL LANGUAGE PROCESSING Cvetana Krstev University of Belgrade, Faculty of Philology Information Retrieval and/vs. Natural Language Processing So close yet so far Outline of the talk ◦ Views on Information Retrieval (IR) and Natural Language Processing (NLP) ◦ IR and NLP in Serbia ◦ Language Resources (LT) in the core of NLP ◦ at University of Belgrade (4 representative resources) ◦ LR and NLP for Information Retrieval and Information Extraction (IE) ◦ at University of Belgrade (4 representative applications) Wikipedia ◦ Information retrieval ◦ Information retrieval (IR) is the activity of obtaining information resources relevant to an information need from a collection of information resources. Searches can be based on full-text or other content-based indexing. ◦ Natural Language Processing ◦ Natural language processing is a field of computer science, artificial intelligence, and computational linguistics concerned with the interactions between computers and human (natural) languages. As such, NLP is related to the area of human–computer interaction. Many challenges in NLP involve: natural language understanding, enabling computers to derive meaning from human or natural language input; and others involve natural language generation. Experts ◦ Information Retrieval ◦ As an academic field of study, Information Retrieval (IR) is finding material (usually documents) of an unstructured nature (usually text) that satisfies an information need from within large collection (usually stored on computers). ◦ C. D. Manning, P. Raghavan, H. Schutze, “Introduction to Information Retrieval”, Cambridge University Press, 2008 ◦ Natural Language Processing ◦ The term ‘Natural Language Processing’ (NLP) is normally used to describe the function of software or hardware components in computer system which analyze or synthesize spoken or written language.
    [Show full text]
  • What Is Special About Patent Information Extraction?
    EEKE 2020 - Workshop on Extraction and Evaluation of Knowledge Entities from Scientific Documents What is Special about Patent Information Extraction? Liang Chen† Shuo Xu Weijiao Shang Institute of Scientific and Technical College of Economics and Research Institute of Forestry Information of China Management Policy and Information Beijing, China P.R. Beijing University of Technology Chinese Academy of Forestry [email protected] Beijing, China P.R. Beijing, China P.R. [email protected] [email protected] Zheng Wang Chao Wei Haiyun Xu Institute of Scientific and Technical Institute of Scientific and Technical Chengdu Library and Information Information of China Information of China Center Beijing, China P.R. Beijing, China P.R. Chinese Academy of Sciences [email protected] [email protected] Beijing, China P.R. [email protected] ABSTRACT However, the particularity of patent text enables the performance of general information-extraction tools to reduce greatly. Information extraction is the fundamental technique for text-based Therefore, it is necessary to deeply explore patent information patent analysis in era of big data. However, the specialty of patent extraction, and provide ideas for subsequent research. text enables the performance of general information-extraction Information extraction is a big topic, but it is far from being methods to reduce noticeably. To solve this problem, an in-depth well explored in IP (Intelligence Property) field. To the best of our exploration has to be done for clarify the particularity in patent knowledge, there are only three labeled datasets publicly available information extraction, thus to point out the direction for further in the literature which contain annotations of named entity and research.
    [Show full text]
  • Naturaljava: a Natural Language Interface for Programming in Java
    NaturalJava: A Natural Language Interface for Programming in Java David Price, Ellen Riloff, Joseph Zachary, Brandon Harvey Department of Computer Science University of Utah 50 Central Campus Drive, Room 3190 Salt Lake City, UT 84112 USA +1 801 581 8224 {deprice,riloff,zachary,blharvey}@cs.utah.edu ABSTRACT We have created NaturalJava, a prototype for an intelligent, natural-language-based user interface that allows programmers to NaturalJava is a prototype for an intelligent natural-language- create, modify, and examine Java programs. With our interface, based user interface for creating, modifying, and examining Java programmers describe programs using English sentences and the programs. The interface exploits three subsystems. The system automatically builds and manipulates a Java abstract Sundance natural language processing system accepts English syntax tree (AST) in response. When the user is finished, the sentences as input and uses information extraction techniques to AST is automatically converted into Java source code. The generate case frames representing program construction and evolving Java program is also displayed in a separate window editing directives. A knowledge-based case frame interpreter, during the programming process so that the programmer can see PRISM, uses a decision tree to infer program modification the code as it is being generated. operations from the case frames. A Java abstract syntax tree manager, TreeFace, provides the interface that PRISM uses to The NaturalJava user interface has three components. The first build and navigate the tree representation of an evolving Java component is Sundance, a natural language processing system program. In this paper, we describe the technical details of each that accepts English sentences as input and uses information component, explain the capabilities of the user interface, and extraction techniques to generate case frames representing present examples of NaturalJava in use.
    [Show full text]
  • Named Entity Recognition: Fallacies, Challenges and Opportunities
    Named Entity Recognition: Fallacies, Challenges and Opportunities Mónica Marrero, Julián Urbano, Sonia Sánchez-Cuadrado, Jorge Morato, Juan Miguel Gómez-Berbís University Carlos III of Madrid Department of Computer Science Avda. Universidad, 30 28911 Leganés, Madrid, Spain [email protected] [email protected] [email protected] [email protected] [email protected] Abstract. Named Entity Recognition serves as the basis for many other areas in Information Management. However, it is unclear what the meaning of Named Entity is, and yet there is a general belief that Named Entity Recognition is a solved task. In this paper we analyze the evolution of the field from a theoretical and practical point of view. We argue that the task is actually far from solved and show the consequences for the development and evaluation of tools. We discuss topics for further research with the goal of bringing the task back to the research scenario. Keywords. Named Entity; Named Entity Recognition; Information Extraction; Evaluation methodology, NER tools 1 Introduction Named Entity Recognition (NER) is a task in Information Extraction consisting in identifying and classifying just some types of information elements, called Named Entities (NE). As such it, serves as the basis for many other crucial areas in Information Management, such as Semantic Annotation, Question Answering, Ontology Population and Opinion Mining. The term Named Entity was first used at the 6th Message Understanding Conference (MUC)[1], where it was clear the importance of the semantic identification of people, organizations and localizations, as well as numerical expressions such as time and quantities. Most of the NER tools nowadays keep considering these types of NE originated in MUC, though with important variations.
    [Show full text]
  • Information Extraction in Text Mining Matt Ulinsm Western Washington University
    Western Washington University Western CEDAR Computer Science Graduate and Undergraduate Computer Science Student Scholarship 2008 Information extraction in text mining Matt ulinsM Western Washington University Follow this and additional works at: https://cedar.wwu.edu/computerscience_stupubs Part of the Computer Sciences Commons Recommended Citation Mulins, Matt, "Information extraction in text mining" (2008). Computer Science Graduate and Undergraduate Student Scholarship. 4. https://cedar.wwu.edu/computerscience_stupubs/4 This Research Paper is brought to you for free and open access by the Computer Science at Western CEDAR. It has been accepted for inclusion in Computer Science Graduate and Undergraduate Student Scholarship by an authorized administrator of Western CEDAR. For more information, please contact [email protected]. Information Extraction in Text Mining Matt Mullins Computer Science Department Western Washington University Bellingham, WA [email protected] Information Extraction in Text Mining Matt Mullins Computer Science Department Western Washington University Bellingham, WA [email protected] ABSTRACT Text mining’s goal, simply put, is to derive information from text. Using multitudes of technologies from overlapping fields like Data Mining and Natural Language Processing we can yield knowledge from our text and facilitate other processing. Information Extraction (IE) plays a large part in text mining when we need to extract this data. In this survey we concern ourselves with general methods borrowed from other fields, with lower-level NLP techniques, IE methods, text representation models, and categorization techniques, and with specific implementations of some of these methods. Finally, with our new understanding of the field we can discuss a proposal for a system that combines WordNet, Wikipedia, and extracted definitions and concepts from web pages into a user-friendly search engine designed for topic- specific knowledge.
    [Show full text]
  • Extracting Latent Beliefs and Using Epistemic Reasoning to Tailor a Chatbot
    Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18) Extracting Latent Beliefs and using Epistemic Reasoning to Tailor a Chatbot Amit Sangroya, C. Anantaram, Pratik Saini, Mrinal Rawat TCS Innovation Labs, Tata Consultancy Services Limited, Gurgaon, India (amit.sangroya, c.anantaram, pratik.saini, rawat.mrinal)@tcs.com Abstract helps to understand the customer’s latent emotional beliefs while giving the complaint. Our model then evaluates the be- During dialog with a customer for addressing liefs to tailor the dialog and make it consistent with the set his/her complaint the chatbot may pose questions or of beliefs of the customer. This process then helps drive the observations based on its underlying model. Some- conversation in a meaningful way. times the questions or observations posed may not be relevant given the nature of complaint and the 2 Proposed Framework current set of beliefs that the customer holds. In this paper we present a framework to build conversa- A customer’s complaint description is usually a set of com- tion system that addresses customer complaints in plex natural language sentences describing the problem faced a meaningful manner using domain understanding, by the customer. Sometimes such descriptions contain back- opinion analysis and epistemic reasoning 1. Extrac- ground information, other superfluous information and may tion of latent beliefs assists in performing epistemic also have customer emotions embedded in it. We first cate- reasoning to maintain a meaningful conversation gorize the customer complaint description into a set of cate- with the customer. gories and then extract relevant information. It is then eval- uated by a hand-crafted knowledge base to assert facts about the probable beliefs held by the human agent in the complaint.
    [Show full text]