Search (2 results, page 1 of 1)

0.016496718 = product of:
  0.06598687 = sum of:
    0.06598687 = weight(_text_:search in 612) [ClassicSimilarity], result of:
      0.06598687 = score(doc=612,freq=8.0), product of:
        0.17183559 = queryWeight, product of:
          3.475677 = idf(docFreq=3718, maxDocs=44218)
          0.049439456 = queryNorm
        0.3840117 = fieldWeight in 612, product of:
          2.828427 = tf(freq=8.0), with freq of:
            8.0 = termFreq=8.0
          3.475677 = idf(docFreq=3718, maxDocs=44218)
          0.0390625 = fieldNorm(doc=612)
  0.25 = coord(1/4)

Abstract

While it is easy to identify whether someone has found a piece of information during a search task, it is much harder to measure how much someone has learned during the search process. Searchers who are learning often exhibit exploratory behaviors, and so current research is often focused on improving support for exploratory search. Consequently, we need effective measures of learning to demonstrate better support for exploratory search. Some approaches, such as quizzes, measure recall when learning from a fixed source of information. This research, however, focuses on techniques for measuring open-ended learning, which often involve analyzing handwritten summaries produced by participants after a task. There are two common techniques for analyzing such summaries: (a) counting facts and statements and (b) judging topic coverage. Both of these techniques, however, can be easily confounded by simple variables such as summary length. This article presents a new technique that measures depth of learning within written summaries based on Bloom's taxonomy (B.S. Bloom & M.D. Engelhart, 1956). This technique was generated using grounded theory and is designed to be less susceptible to such confounding variables. Together, these three categories of measure were compared by applying them to a large collection of written summaries produced in a task-based study, and our results provide insights into each of their strengths and weaknesses. Both fact-to-statement ratio and our own measure of depth of learning were effective while being less affected by confounding variables. Recommendations and clear areas of future work are provided to help continued research into supporting sensemaking and learning.

0.013997929 = product of:
  0.055991717 = sum of:
    0.055991717 = weight(_text_:search in 3788) [ClassicSimilarity], result of:
      0.055991717 = score(doc=3788,freq=4.0), product of:
        0.17183559 = queryWeight, product of:
          3.475677 = idf(docFreq=3718, maxDocs=44218)
          0.049439456 = queryNorm
        0.3258447 = fieldWeight in 3788, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          3.475677 = idf(docFreq=3718, maxDocs=44218)
          0.046875 = fieldNorm(doc=3788)
  0.25 = coord(1/4)

Abstract

We address the challenge of extracting query biased audio summaries from podcasts to support users in making relevance decisions in spoken document search via an audio-only communication channel. We performed a crowdsourced experiment that demonstrates that transcripts of spoken documents created using Automated Speech Recognition (ASR), even with significant errors, are effective sources of document summaries or "snippets" for supporting users in making relevance judgments against a query. In particular, the results show that summaries generated from ASR transcripts are comparable, in utility and user-judged preference, to spoken summaries generated from error-free manual transcripts of the same collection. We also observed that content-based audio summaries are at least as preferred as synthesized summaries obtained from manually curated metadata, such as title and description. We describe a methodology for constructing a new test collection, which we have made publicly available.