Search (8 results, page 1 of 1)

0.01586826 = product of:
  0.03173652 = sum of:
    0.03173652 = sum of:
      0.0061926404 = weight(_text_:a in 2640) [ClassicSimilarity], result of:
        0.0061926404 = score(doc=2640,freq=10.0), product of:
          0.043477926 = queryWeight, product of:
            1.153047 = idf(docFreq=37942, maxDocs=44218)
            0.037706986 = queryNorm
          0.14243183 = fieldWeight in 2640, product of:
            3.1622777 = tf(freq=10.0), with freq of:
              10.0 = termFreq=10.0
            1.153047 = idf(docFreq=37942, maxDocs=44218)
            0.0390625 = fieldNorm(doc=2640)
      0.02554388 = weight(_text_:22 in 2640) [ClassicSimilarity], result of:
        0.02554388 = score(doc=2640,freq=2.0), product of:
          0.13204344 = queryWeight, product of:
            3.5018296 = idf(docFreq=3622, maxDocs=44218)
            0.037706986 = queryNorm
          0.19345059 = fieldWeight in 2640, product of:
            1.4142135 = tf(freq=2.0), with freq of:
              2.0 = termFreq=2.0
            3.5018296 = idf(docFreq=3622, maxDocs=44218)
            0.0390625 = fieldNorm(doc=2640)
  0.5 = coord(1/2)

Abstract

We propose a tag-based framework that simulates human abstractors' ability to select significant sentences based on key concepts in a sentence as well as the semantic relations between key concepts to create generic summaries of transcribed lecture videos. The proposed extractive summarization method uses tags (viewer- and author-assigned terms) as key concepts. Our method employs Flickr tag clusters and WordNet synonyms to expand tags and detect the semantic relations between tags. This method helps select sentences that have a greater number of semantically related key concepts. To investigate the effectiveness and uniqueness of the proposed method, we compare it with an existing technique, latent semantic analysis (LSA), using intrinsic and extrinsic evaluations. The results of intrinsic evaluation show that the tag-based method is as or more effective than the LSA method. We also observe that in the extrinsic evaluation, the grand mean accuracy score of the tag-based method is higher than that of the LSA method, with a statistically significant difference. Elaborating on our results, we discuss the theoretical and practical implications of our findings for speech video summarization and retrieval.

Date

22. 1.2016 12:29:41

Type

a

0.0023983994 = product of:
  0.004796799 = sum of:
    0.004796799 = product of:
      0.009593598 = sum of:
        0.009593598 = weight(_text_:a in 5358) [ClassicSimilarity], result of:
          0.009593598 = score(doc=5358,freq=24.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.22065444 = fieldWeight in 5358, product of:
              4.8989797 = tf(freq=24.0), with freq of:
                24.0 = termFreq=24.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.0390625 = fieldNorm(doc=5358)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

We propose and evaluate a video summarization method based on a topic relevance model, a maximal marginal relevance (MMR), and discriminant analysis to generate a semantically meaningful video skim. The topic relevance model uses event-related potential (ERP) components to describe the process of topic relevance judgment. More specifically, the topic relevance model indicates that N400 and P600, which have been successfully applied to the mismatch process of a stimulus and the discourse-internal reorganization and integration process of a stimulus, respectively, are used for the topic mismatch process of a topic-irrelevant video shot and the topic formation process of a topic-relevant video shot. To evaluate our proposed ERP/MMR-based method, we compared the video skims generated by the ERP/MMR-based, ERP-based, and shot boundary detection (SBD) methods with ground truth skims. The results showed that at a significance level of 0.05, the ROUGE-1 scores of the ERP/MMR method are statistically higher than those of the SBD method, and the diversity scores of the ERP/MMR method are statistically higher than those of the ERP method. This study suggested that the proposed method may be applied to the construction of a video skim without operational intervention, such as the insertion of a black screen between video shots.

Type

a

0.0022962927 = product of:
  0.0045925854 = sum of:
    0.0045925854 = product of:
      0.009185171 = sum of:
        0.009185171 = weight(_text_:a in 1882) [ClassicSimilarity], result of:
          0.009185171 = score(doc=1882,freq=22.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.21126054 = fieldWeight in 1882, product of:
              4.690416 = tf(freq=22.0), with freq of:
                22.0 = termFreq=22.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.0390625 = fieldNorm(doc=1882)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

This study develops evaluation indicators for a consortium of Korean institutional repositories called dCollection and validates the indicators against actual data from the participants of this consortium. The literature review reveals a conceptual framework for institutional repository (IR) evaluation with four categories and 19 items. In developing the initial framework, equal amounts of emphasis are put on the assessments of procedural achievement and actual performance to pinpoint the procedural weaknesses of each IR and to determine its customized solution. A Delphi method of three rounds with the help of IR librarians reveals a converging tendency pertaining to the measures of importance ascribed to the categories and items. Through a focus-group interview with middle- to top-level managers, 39 indicators derived from 19 items are identified as possessing relevancy, measurability, data availability, and differentiability. Validation of evaluation indicators employs actual evaluation data from 32 university IRs. Factor analysis shows a simpler structural pattern containing 12 factors than that of the structural pattern of the conceptual framework that contains 19 items. Correlation analysis using the factor scores identifies six key factors: Registration Rate, Archiving, Resource Allocation, System Performance, Multifunctionality, and Use Rate. The results from regression analyses suggest that two different approaches can be employed to promote the Use Rate factor. In the content-oriented approach, the Registration Rate factor is crucial while in the policy-oriented approach the Archiving factor assumes this role; however, the System Performance factor plays a mediating role for the key factors, thus forming a contingency for either approach.

Type

a

0.0020770747 = product of:
  0.0041541494 = sum of:
    0.0041541494 = product of:
      0.008308299 = sum of:
        0.008308299 = weight(_text_:a in 4685) [ClassicSimilarity], result of:
          0.008308299 = score(doc=4685,freq=18.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.19109234 = fieldWeight in 4685, product of:
              4.2426405 = tf(freq=18.0), with freq of:
                18.0 = termFreq=18.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.0390625 = fieldNorm(doc=4685)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

Our aim was to develop an event-related potential (ERP)-based method to construct a video skim consisting of key shots to bridge the semantic gap between the topic inferred from a whole video and that from its summary. Mayer's cognitive model was examined, wherein the topic integration process of a user evoked by a visual stimulus can be associated with long-latency ERP components. We determined that long-latency ERP components are suitable for measuring a user's neuronal response through a literature review. We hypothesized that N300 is specific to the categorization of all shots regardless of topic relevance, N400 is specific for the semantic mismatching process for topic-irrelevant shots, and P600 is specific for the context updating process for topic-relevant shots. In our experiment, the N400 component led to more negative ERP signals in response to topic-irrelevant shots than to topic-relevant shots and showed a fronto-central scalp pattern. P600 elicited more positive ERP signals for topic-relevant shots than for topic-irrelevant shots and showed a fronto-central scalp pattern. We used discriminant and artificial neural network (ANN) analyses to decode video shot relevance and observed that the ANN produced particularly high success rates: 91.3% from the training set and 100% from the test set.

Type

a

0.0019582848 = product of:
  0.0039165695 = sum of:
    0.0039165695 = product of:
      0.007833139 = sum of:
        0.007833139 = weight(_text_:a in 3448) [ClassicSimilarity], result of:
          0.007833139 = score(doc=3448,freq=16.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.18016359 = fieldWeight in 3448, product of:
              4.0 = tf(freq=16.0), with freq of:
                16.0 = termFreq=16.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.0390625 = fieldNorm(doc=3448)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

Two key problems in developing a storyboard are (a) the extraction of video key frames and (b) the display of the storyboard. On the basis of our findings from a preliminary study as well as the results of previous studies on the computerized extraction of key frames and human recognition of images and videos, we propose an algorithm for the extraction of key frames and the structural display of a storyboard. In order to evaluate the proposed algorithm, we conducted an experiment, the results of which suggest that participants produce better summaries of the given videos when they view storyboards that are composed of key frames extracted using the proposed algorithmic method. This finding held, regardless of whether the display pattern used was sequential or structural. In contrast, the experimental results suggest that in the case of employing a mechanical method, the use of a structural display pattern yields greater performance in terms of participants' ability to summarize the given videos. Elaborating on our results, we discuss the practical implications of our findings for video summarization and retrieval.

Type

a

0.001938603 = product of:
  0.003877206 = sum of:
    0.003877206 = product of:
      0.007754412 = sum of:
        0.007754412 = weight(_text_:a in 2181) [ClassicSimilarity], result of:
          0.007754412 = score(doc=2181,freq=2.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.17835285 = fieldWeight in 2181, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.109375 = fieldNorm(doc=2181)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Type

a

0.0016616598 = product of:
  0.0033233196 = sum of:
    0.0033233196 = product of:
      0.006646639 = sum of:
        0.006646639 = weight(_text_:a in 4229) [ClassicSimilarity], result of:
          0.006646639 = score(doc=4229,freq=8.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.15287387 = fieldWeight in 4229, product of:
              2.828427 = tf(freq=8.0), with freq of:
                8.0 = termFreq=8.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046875 = fieldNorm(doc=4229)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

The design and implementation of an ontology-based Web retrieval (ONTOWEB) system is described; ONTOWEB allows the semantic search of the Web resources of international organizations such as the World Bank and the Organisation for Economic Co-operation and Development (OECD). A firm's knowledge management project is introduced first, followed by a description of the ONTOWEB system's design, implementation, and evaluation. The ONTOWEB system has two components: databases and an ontology-based search engine. The ontology-based search engine is a tool used to query the information that has been loaded into the database. Last, to evaluate the system, an experiment was conducted to compare the performance of the proposed system with that of Internet search engines in terms of relevance and search time. This study shows that ontologies can be used not only to improve precision, but also to reduce the search time. Because of the expense of annotating resources, the domains that contain the most valuable knowledge, such as the medical and the business sectors, are prime areas for future ontology applications.

Type

a

0.0013847164 = product of:
  0.0027694327 = sum of:
    0.0027694327 = product of:
      0.0055388655 = sum of:
        0.0055388655 = weight(_text_:a in 4350) [ClassicSimilarity], result of:
          0.0055388655 = score(doc=4350,freq=8.0), product of:
            0.043477926 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.037706986 = queryNorm
            0.12739488 = fieldWeight in 4350, product of:
              2.828427 = tf(freq=8.0), with freq of:
                8.0 = termFreq=8.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.0390625 = fieldNorm(doc=4350)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

This study investigated the effectiveness of query expansion using synonymous and co-occurrence tags in users' video searches as well as the effect of visual storyboard surrogates on users' relevance judgments when browsing videos. To do so, we designed a structured folksonomy-based system in which tag queries can be expanded via synonyms or co-occurrence words, based on the use of WordNet 2.1 synonyms and Flickr's related tags. To evaluate the structured folksonomy-based system, we conducted an experiment, the results of which suggest that the mean recall rate in the structured folksonomy-based system is statistically higher than that in a tag-based system without query expansion; however, the mean precision rate in the structured folksonomy-based system is not statistically higher than that in the tag-based system. Next, we compared the precision rates of the proposed system with storyboards (SB), in which SB and text metadata are shown to users when they browse video search results, with those of the proposed system without SB, in which only text metadata are shown. Our result showed that browsing only text surrogates-including tags without multimedia surrogates-is not sufficient for users' relevance judgments.

Type

a