We are living in the data-centric era; data is omnipresence in either structured or
unstructured forms. This data is generated through social media, blogs, lab simulations;
sensors etc. for user related operations. Due to this big data occurrence, acquisition of
relevant information, becomes a challenging task. The users of such systems are oblivious
about the data content and its semantics and often uncertain of their real information
needs. Here, evolving search needs are additional challenges. A user often struggle in
constructing a query that retrieves relevant information, and hence ended up with a short
close ill-phrased navigational query. In this setting, the traditional search systems do
information search inefficiently and thus becomes a daunting task. Hence, a data
exploration (DE) technique is evolved. The data exploration or say interactive data
exploration(IDE) guides user throughout his search session by retrieving valuable data
objects on-the-fly.
Data exploration is iterative, multi-disciplinary and opportunistic user behavior of
information search. For effective data exploration the key challenges are (i) vaguely
articulated user’s information needs (ii) utilization of available data space for exploration
(iii) identification of shift in the user’s intends (iv) information overload and many more.
Therefore, instead of this traditional search, we need data exploration mechanism in which
a naïve user walks through the database and stops when satisfactory information is met.
During this, a user iteratively transforms his search request in order to gain relevant
information; morphing is a historic approach for the generation of a various transformation
of input.
A proximity-based approach for data exploration and query reformulation to aide
exploratory search is the key objective of dissertation. In proposed approach, data objects
are retrieves along with additional information via a small transformation of user’s initial
query, called as ‘Query Morphs’. The dissertation’s contribution is: (i) to design an
exploratory system that supports data exploration and inherits query reformulation. (ii) to
assist user in his exploratory query formulation task via additional results. (iii) to aide, in
generation of query variants (iv) to enable user, for steering towards the region-of-interest.
(v) Minimize user effort and improve search precession. The assessment of proposed
strategy indicates significant improvement in data exploration and query reformulation
effort of the user and DE system.
Figures - uploaded by
Vikram SinghAuthor contentAll figure content in this area was uploaded by Vikram Singh
Content may be subject to copyright.