Abstract | ||
---|---|---|
In this paper, we present a robust passage retrieval algorithm to extend the conventional text question answering (Q/A) to videos. Users interact with our videoQ/A system through natural language queries, while the top-ranked passage fragments with associated video clips are returned as answers. We compare our method with five of the high-performance ranking algorithms that are portable to different languages and domains. The experiments were evaluated with 75.3 h of Chinese videos and 253 questions. The experimental results showed that our method outperformed the second best retrieval model (language models) in relatively 1.43% in mean reciprocal rank (MRR) score and 11.36% when employing a Chinese word segmentation tool. By adopting the initial retrieval results from the retrieval models, our method yields an improvement of at least 5.94% improvement in MRR score. This makes it very attractive for the Asia-like languages since the use of a well-developed word tokenizer is unnecessary. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/TCSVT.2008.2002831 | IEEE Trans. Circuits Syst. Video Techn. |
Keywords | Field | DocType |
mrr score,different language,video clips,text question answering,video question answering (videoq/a),mean reciprocal rank,initial retrieval result,language models,chinese videos,video question answering,best retrieval model,method yield,asia-like languages,chinese word segmentation tool,word tokenizer,chinese video,multimedia retrieval,robust passage retrieval algorithm,asia-like language,natural language processing,question answering (q/a),video retrieval,retrieval model,query processing,natural languages,data mining,robustness,image retrieval,search engines,indexing terms,language model,natural language,information retrieval,question answering | Learning to rank,Question answering,Computer science,Image retrieval,Speech recognition,Text segmentation,Mean reciprocal rank,Natural language,Natural language processing,Artificial intelligence,Lexical analysis,Language model | Journal |
Volume | Issue | ISSN |
18 | 10 | 1051-8215 |
Citations | PageRank | References |
11 | 0.58 | 40 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yu-Chieh Wu | 1 | 247 | 23.16 |
Jie-Chi Yang | 2 | 350 | 43.91 |