Utilizing Entities for an Enhanced Search Experience

Abstract

Over the past decade, search engines have not only improved the quality of search results, but also evolved in how they interact with users. Modern search engines provide assistance to users throughout the entire search process, from formulating their information needs to presenting results and recommending additional content. This chapter presents a selection of topics, where entities are utilized with the overall aim of improving users' search experiences. First, in Sect. 9.1, we discuss techniques for assisting users with articulating their information needs. These include query assistance services, such as query autocompletion and query suggestions, and specialized query building interfaces. Next, in Sect. 9.2, we turn to the question of result presentation. In conventional document retrieval, the standard way of serving results is to display a snippet for each document, consisting of its title and a short summary. This summary is automatically extracted from the document with the aim of explaining why that particular document is relevant to the query at hand. Moving from documents to entities as the unit of retrieval, the question we need to ask is: How can one generate dynamic summaries of entities when displaying them as search results? Finally, in Sect. 9.3, we describe entity recommendation methods that present users with contextual suggestions, encourage exploration, and allow for serendipitous discoveries. We study the related entity retrieval problem in different flavors, depending on what kind of input is available to the recommendation engine. Furthermore, we address the question of explaining the relationship in natural language between entities presented to the user. We refer to Table 9.1 for the notation used in this chapter. 9.1 Query Assistance Chapter 7 dealt with query understanding from the machine's point of view. In this section, we bring the user's perspective to the forefront. How can a semantic search

Graph-Based Entity-Oriented Search: A Unified Framework in Information Retrieval

“…While the first is focused on unstructured data (documents), the second largely relies on structured data (entities and relations). Moreover, entity-oriented search also incorporates ad hoc document retrieval, as long as it leverages entities [2,Ch. 8].…”

Health Cards for Consumer Health Search

“…Health cards more generally, and outside the health domain, can be referred to as entity cards or information cards [29]. An entity card presents a rich and coherent set of information about a specic entity; this commonly includes the entity's name and type, a textual summary, a factual summary, key features, relationships, and links to related entities [3,29]. Entity cards are now an integral part of the SERP in commercial search engines like Google, Bing, and Yandex.…”

“…Entity cards are now an integral part of the SERP in commercial search engines like Google, Bing, and Yandex. Studies show their use improves user engagement, attracts user attention, and enhances user experience [3,6].…”

“…After consenting to participate, each participant was given a set of instructions presenting the elements of the interface and rules for the collection of evidence to answer the scenarios. Next, a demographic questionnaire collected information on the participant's age group (grouped by ten-years intervals 2 ), highest level of completed education, English prociency 3 , and the frequency of use of generalpurpose search engines. We used the responses to determine the participant's eligibility, as described in Section 3.10.…”

“…Following the guidelines for age-group data anonymisation from the Australian Bureau of Statistics 3. We veried participants English prociency by checking whether they: (1) speak English as rst language, or (2) achieved IELTS overall test score of at least 5.0 with a score of at least 4.5 in each of the four test components.…”