An increasing number of databases have become web accessible through HTML form-based search interfaces. The data units returned from the underlying database are usually encoded into the result pages dynamically for human browsing. For the encoded data units to be machine processable, which is essential for many applications such as deep web data collection and Internet comparison shopping, they need to be extracted out and assigned meaningful labels. In this paper, we present an automatic annotation approach that first aligns the data units on a result page into different groups such that the data in the same group have the same semantic. Then, for each group we annotate it from different aspects and aggregate the different annotations to predict a final annotation label for it. An annotation wrapper for the search site is automatically constructed and can be used to annotate new result pages from the same web database. Our experiments indicate that the proposed approach is highly effective.

An increasing number of databases have become Web accessible through HTML form-based search interfaces. The data units returned from the underlying database are usually encoded into the result pages dynamically for human browsing. For the encoded data units to be machine processable, which is essential for many applications such as deep Web data collection and comparison shopping, they need to be extracted out and assigned meaningful labels. In this paper, we present a multi-annotator approach that first aligns the data units into different groups such that the data in the same group have the same semantics. Then for each group, we annotate it from different aspects and aggregate the different annotations to predict a final annotation label. An annotation wrapper for the search site is automatically constructed and can be used to annotate new result pages from the same site. Our experiments indicate that the proposed approach is highly effective.

Metasearch engine, Comparison-shopping and Deep Web crawling applications need to extract search result records enwrapped in result pages returned from search engines in response to user queries. The search result records from a given search engine are usually formatted based on a template. Precisely identifying this template can greatly help extract and annotate the data units within each record correctly. In this paper, we propose a graph model to represent record template and develop a domain independent statistical method to automatically mine the record template for any search engine using sample search result records. Our approach can identify both template tags (HTML tags) and template texts (non-tag texts), and it also explicitly addresses the mismatches between the tag structures and the data structures of search result records. Our experimental results indicate that this approach is very effective.

AllInOneNews is the largest news metasearch engine in the world, connecting to over 1,000 news sites over 150 countries. Implementing a large-scale metasearch engine like AllInOneNews needs to overcome unique challenges not faced by building small metasearch engines such as developing highly scalable search engine selection techniques. In this paper, we discuss these unique challenges and our solutions to these challenges. We also discuss some novel features of AllInOneNews such as highly automated solution and semantic query match. This paper also reports the results of a comparative evaluation of three commercial news search systems, one search engine - Google News and two metasearch engines - Mamma News and AllInOneNews. Several measures such as effectiveness, diversity and time-sensitivity are used to perform the comparison. Another contribution of this paper is that we introduce a novel scheme to compare multiple news search systems in a combined measure that takes both relevance and time-sensitivity of retrieved information into consideration.

In this paper, we describe MySearchView, a system for assembling search engines into metasearch engines. With this system, any user can create a metasearch engine by simply letting the system know the URLs of the search engines the user wants to be included and the metasearch engine will be built fully automatically. In this paper, the main steps of building metasearch engines will be sketched. We will also outline our plan to demonstrate all the features of MySearchView.

When a query is submitted to a search engine, the search engine returns a dynamically generated result page containing the result records, each of which usually consists of a link to and/or snippet of a retrieved Web page. In addition, such a result page often also contains information irrelevant to the query, such as information related to the hosting site of the search engine and advertisements. In this paper, we present a technique for automatically producing wrappers that can be used to extract search result records from dynamically generated result pages returned by search engines. Automatic search result record extraction is very important for many applications that need to interact with search engines such as automatic construction and maintenance of metasearch engines and deep Web crawling. The novel aspect of the proposed technique is that it utilizes both the visual content features on the result page as displayed on a browser and the HTML tag structures of the HTML source file of the result page. Experimental results indicate that this technique can achieve very high extraction accuracy.