Organizing and searching the world wide web of facts -- step two

Abstract

As part of a large effort to acquire large repositories of facts from unstructured text on the Web, a seed-based framework for textual information extraction allows for weakly supervised extraction of class attributes (e.g., side effects and generic equivalent for drugs) from anonymized query logs. The extraction is guided by a small set of seed attributes, without any need for handcrafted extraction patterns or further domain-specific knowledge. The attributes of classes pertaining to various domains of interest to Web search users have accuracy levels significantly exceeding current state of the art. Inherently noisy search queries are shown to be a highly valuable, albeit unexplored, resource for Web-based information extraction, in particular for the task of class attribute extraction.

Queries as a Source of Lexicalized Commonsense Knowledge

“…Each query is available independently from other queries, and is accompanied by its frequency of occurrence in Target Classes: Table 2 shows the set of 40 target classes for evaluating the extracted facts. Similar evaluation strategies were followed in previous work (Paşca, 2007). As illustrated earlier in Figure 1, a target class consists in a small set of phrase descriptors.…”

Improved Method for Relation Adaptation to Extract the New Relations Efficiently from Relational Mapping

“…In [3], authors M. Pasca et al presented the bootstrapping methods to relation extraction are attractive because they require markedly fewer training instances than supervised approaches do. Bootstrapping methods are initialized with a few instances of the target relation.…”

Automatic set instance extraction using the web

“…As mentioned earlier, their system requires the name of the semantic set and an additional seed instance. Pasca [7,6] illustrated an set expansion approach that extracts instances from Web search queries given a set of input seed instances, which is similar in flavor to our system SEAL but different from the task addressed in this paper: the user provides no seeds, but instead provides the name of the set being expanded.…”