Content-Determined Web Page Segmentation and Navigation for Mobile Web Searching

Nowadays the usage of mobile phones is widely spread in our lifestyle; we use cell phones as a camera, a radio, a music player, and even as a web browser. Since most web pages are created for desktop computers, navigating through web pages is highly fatigued. Hence, there is a great interest in computer science to adopt such pages with rich content into small screens of our mobile devices. On the other hand, every web page has got many different parts that do not have the equal importance to the end user. Consequently, the authors propose a mechanism to identify the most useful part of a web page to a user regarding his or her search query while the information loss is avoided. The challenge here comes from the fact that long web contents cannot be easily displayed in both vertical and horizontal ways.

Horizontal Web Searching and Navigational Resource Identification

Generally speaking, horizontal search engines are meant to deal with general web queries. In the context of this chapter, the authors investigated the act of navigational resource identification in the light of horizontal web searching. State-of-the-art navigational resource identification is reluctant to the distinct characteristics of the navigational queries and specific users' treatments toward different searching tasks. Consequently, in this chapter, the authors discussed a new mechanism for navigational resource identification according to previous findings.

Vertical Result Page Generation for Academic Web Searching

Vertical search engines are meant for answering a user's web query within a specific domain such as news, media, and academic web searching. One main difference between vertical and horizontal web searching is that in vertical web searching, unlike horizontal web searching, a subset of entire web is engaged. The chapter investigates the state-of-the-art in academic web searching and points out shortcomings in this particular domain. Lastly, the authors aimed to propose a summary-based recommender to respond to a user's query by retrieving and ranking them according to their similarity merits on the basis of papers' summaries. Results of the evaluations revealed the fact that the proposed framework has outperformed the state-of-the-art in different metrics such as unanimous ranks and F1 measures.

Enhancing Academic Recommendation Regarding Common Coauthors' Publication Records

In this chapter, the authors investigated the feasibility of any improvement in paper recommendation by recommending similar papers to an input paper from the publication record of the first author. Although there are numerous approaches for recommending academic papers, they did not consider intellectually recommending papers based on the publication record of common coauthors. Consequently, they are motivated to introduce a remedy for this shortcoming by recommending scholarly papers based on similarity of textual references to visual features which considers the similarity of text fragments of one's publication record to any of their visual features (i.e., tables and figures). Based on the results of evaluation, the proposed enhancement will increase the mean precision, recall, and accordingly, the F-measure. In addition, it increases the position of the relevant papers in the returned list of documents.

Ambiguity and Clarification

By the time web engines were developed, the number of queries prompted by users had grown exponentially. This fast growth shows the high demand of users from web search engines. This high demand made search engines responsible for the users' satisfaction during a search session. One way to improve a user's satisfaction is to visualize search engine result page (SERP). Recent studies for meeting this aim focused on a whole page thumbnail for assisting users to remember recently visited web pages. This chapter explores how a specific visual content of a page can allow users to distinguish between a useful and worthless page within results in SERP especially in an ambiguous search task.

A Comparative Web Usage Study of Users in Navigational Web Searching

This chapter highlighted the differences of users' behaviors between native English-speaking users and Chinese users as the biggest example of non-native English-speaking users. To do as such, the author of this chapter began by discussing the background of earlier web usage studies followed by a literature review on comparative studies that are on the basis of users with different language preferences. Afterwards, since earlier web log analyses are based on web transactions collected from mainly native users, the author of this chapter investigated the feasibility of generalization of former findings for navigational searching to the rest of the users by comparing two web log transactions from two groups of users with different localities in respect to the state-of-the-art in web searching.

Navigational Searching and User Treatments

In this chapter, the authors presented particularity of web searching in the context of navigational searching based on previous studies. Consequently, they divided this chapter into two parts. The first part of this chapter belongs to earlier works that examined eye-tracking studies to investigate distinctiveness between different searching tasks while the second part belongs to the discussion on the structure and nature of each searching task based on the results of earlier web usage studies.