Cobot is a new intelligent agent platform that connects users through real-time and o -line conversations about their health and medical issues. Intelligent web based information agents (conversational/community bots) participate in each conversation providing highly-relevant real-time informational recommendations and connecting people with relevant conversations and other community members. Cobot provides an innovative approach to facilitate easier information access allowing users to exchange information through a natural language conversational approach. Conversational Search(CS) is an interactive and collaborative information nding interaction. The participants in this interaction engage in social conversations aided with an intelligent information agent (Cobot) that provides contextually relevant factual, web search and social search recommendations. Cobot aims to help users make faster and more informed search and discovery. It also helps the agent learn about conversations with interactions and social feedback to make better recommendations. Cobot leverages the social discovery process by integrating web information retrieval along with the social interactions and recommendations.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
Copyright 200X ACM X-XXXXX-XX-X/XX/XX ...$10.00.
This paper introduces a novel Conversational Search and
Recommendation system that involves nding relevant
information based on social interactions and feedback along
with augmented agent based recommendations. People in
social groups can provide solutions (answers to questions)[
We are building a system called Cobot1 to address some of these challenges. Cobot introduces a conversational environment that provides social search through conversations integrated with intelligent semantic meta-search from the web. Users want to simplify their experience when performing an information nding task. Conversational Search is about letting users collaboratively search and nd in natural language, leaving the task of user intent comprehension on the system. The participating agent interacts with users proving recommendations that the users can accept, reject, like, dislike or suggest. 2.
Cobot is an intelligent agent platform that connects users through real-time and o ine conversations. Cobot lives in a community, has a limited understanding of domains through ontologies and brings relevant information to the users by participating in the conversations. Cobot's 'conversation engine' monitors user conversations with other users in the community and provides/receives recommendations (links and snippets) based on the conversation to the participants. Cobot's `community engine' models conversations to capture user-user and user-information interactions.
Design Goals 1. Near real time conversational agent 2. Personalized as well as generic recommendations 3. Agent learns with interaction 4. Uses a structured internal organization of content 1We use the term Cobot for Cobot system as well as Cobot agent interchangeably 5. Dynamically connects conversations to the right set of people for participation 6. Helps the user talk about health issues. (Real time conversations) 7. In the real time conversational process, provides recommendations. (`who to talk to', `what to look at') In the following sections, we brie y describe the features of the Cobot system. 2.1
Cobot provides a conversational interface that combines semantic language understanding and real time collaborative environment for information retrieval with contextually relevant recommendations. Cobot helps people nd information faster with the aim for nding useful responses for completing the search intent. The conversational interface allows for much more interactivity than one-shot search style interfaces, which aids usability and improves intent understanding. For example, Cobot recommends links and snippets from relevant articles on the internet. It also makes social recommendations to connect contextually relevant users to the conversation.
The approach we have taken to address CS problems is by developing dynamic data structures that model it. We call this structure the \Socio-Semantic Model" - these conversation nets maintain in memory models of the conversation, participants, participants' immediate social connections, concepts, relationships and information ow. 2.2
The Socio-Semantic Conversation Model is a dynamic memory data structure based on principles of experience based agent architecture.[10] It supports interleaved retrieval of information by applying di erent memory retrieval algorithms. The model maintains the user's social graph, the conversation graph with the extracted semantic net for the conversation.
Some essential properties of the model are as follows: The model is socially aware of the participant and his social network's availability (to aid with Cohort Matching) The model provides bi-directional recommendation and feedback. (Both agent and the participant can add recommendations) The model understands limited domain terminology and is able to nd semantic relationships amongst concepts extracted from conversations.
The model is aware of user's pro le (such as interests and ratings) for the agent to be able to use that information.
The Socio-Semantic Model aims to provide storage and memory based retrieval for dynamic representation, update and reuse of users' knowledge and experiences. Figure 1 depicts the user-centric domain information modeling approach to jointly model the information context from users' perspective. 2.3
Identifying relevant documents for a particular user's need without extensive search, in conversational manner is the key objective for precise search. The right search queries need to be gured out with situation assessment from the conversational snippets. It is not desirable to return dozens or hundreds of remotely relevant results, even if some of them will be highly relevant. The aim is to retrieve successive recommendations that try to address the search problem precisely. Cobot uses di erent shallow semantic parsing techniques for operationalizing a user's intent into computational form, dispatching to multiple, heterogeneous services, gathering and integrating results, and presenting them back to the user as a set of solutions to their request. 2.4
Communities are made up of users who are grouped by di erent information needs into dynamic cohorts. These online communities, through e ective sharing and collaboration, increase the utility of systems and help solve individual problems more e ectively. Cobot allows for connecting two or more individuals to an online conversation based on the topic and context of conversation, mutual interests, and what they want to talk about at that time. The system allows any individual to nd/join that conversation. 2.5
Filtering and recommendation are crucial in collaborative systems enabling users to navigate an ever-growing deluge of information more e ectively. Cobot's recommendation engine delivers quality information delivered through lters achieved from semantic and contextual understanding of text along with captured users' interests. It uses various personalization techniques such as collaborative ltering on conversations and other entities in context. Natural language processing techniques are used to enhance the content based recommendations.[8] 3.
This paper proposes a collaborative system for conversational search and recommendations. We are hypothesizing