2.2 Question Recommendation
should I buy more memory or s-
Wu et al. (2008) presented an incremental auto-
torage space? ...
matic question recommendation framework based
Q Body I edit pictures and videos so I
on probabilistic latent semantic analysis. Question
need them to work quickly. Any
recommendation in their work considered both the
advice?
users’ interests and feedback. Duan et al. (2008)
Answer ... If you are running out of s-
made use of a tree-cut model to represent question-
pace on your hard drive, then
s as graphs of topic terms. Questions were recom-
... to boost your computer speed
mended based on this topic graph. The recommend-
usually requires more RAM ...
ed questions can provide different aspects around the
topic of the query question.
Table 1: Yahoo! Answers question exampleThe above question search and recommendation
research provide different ways to retrieve question-
s from large archives of question answering data.
lows. In section 2, we briefly describe the related
However, none of them considers the similarity or
work on question search and recommendation. Sec-
diversity between questions by exploring their infor-
tion 3 addresses in detail how we measure the sim-
mation needs.
ilarity between short texts. Section 4 describes two
models for information need prediction that we use
3 Short Text Similarity Measures
for the experiment. Section 5 tests the performance
In question retrieval systems accurate similarity
of the proposed models for the task of question rec-
measures between documents are crucial. Most tra-
ommendation. Section 7 is the conclusion of this
ditional techniques for measuring the similarity be-
paper.
tween two documents mainly focus on comparing
word co-occurrences. The methods employing this
2 Related Work
strategy for documents can usually achieve good re-
sults, because they may share more common words
than short text snippets. However the state-of-the-
Bạn đang xem 2. - BÁO CÁO KHOA HỌC: "IMPROVING QUESTION RECOMMENDATION BY EXPLOITING INFORMATION NEED" PPTX