22 nd User Modeling, Adaptation and Personalization (UMAP 2014) Time-Sensitive User Profile for Optimizing Search Personalization Ameni Kacem, Mohand Boughanem,

22nd User Modeling, Adaptation and Personalization (UMAP 2014)

Time-Sensitive User Profile forOptimizing Search Personalization

Ameni Kacem, Mohand Boughanem, Rim Faiz

Introduction

Related Work

Time-Based User Profile

Conclusion

Experiments and Results

Outline

3

Introduction

4Context (1)

Personalization: search results adapted to the user’s information needs and inetrests.

Time integration

when the words appear

Time to discern short- term and long –term uer

profiles

Context (2)

• Interactions extracted from the current sesssion

• No long-term interests

Short-term

• Old interests• No consideration of the actual user needs

Long-term

5

6Problem Description (1)

Weight the profiles terms: both the freshness and the frequency

Unify both the recent and persistent interests.

User interests evolution over time

A time sensitive user profile (older frequent terms should not outperform current and not frequent terms).

Problem Description (2)7

How do temporal dynamics affect the quality of user models in the context of personalized search ?

How short-term (recent) profile and long-term (persistent) profile interact ?

How each of the profiles may be used in separation or unified ?

Related Work

9User Profiling

User profiling

Information about the user from different sources

Multiple representations:

Vector• Weighted

keywords

Categories• Open

Directory Project

Semantic network• Concepts

10Short-term user profile

Short-term: the interests and needs of users related to activities of the current search session.

Daoud et al., 2009; Zemirli, 2008: the short-term user profile is all the interactions and interests related to a single information need.

Dumais et al., 2003; Shen et al. 1999: it represents multiple interests emerged in a single time slot.

11Long-term user profile

Long-term: The use of specific information: education level, general interests, user query history and past user clickthrough information.

Teevan et al. (2005): rich long-term user models based on desktop search activities to improve ranking.

Tan et al. (2006): long-term language model-based representations of users’ interests based on queries, documents and clicks.

12Recent Similar Works

Paper Approach

Bennett et al. (2012)

The first study to assess how short-term and long-term behaviors relate, and how each may be used in isolation or in combination.

Abel et al. (2013)

Different strategies for mining user interest profiles from microblogging activities ranging such as strategies that adapt to temporal patterns that can be observed in the microblogging behavior.

13Temporal User Profile

The user profile can reflect both the recurrent (persistent) and the current (recent) interests but with different scales based on freshness.

Users who are not very active

• The short-term profile can eliminate relevant results which are more related to their personal interests.

For users who are very active

• The aggregation of recent activities without ignoring the old interests would be very interesting.

Our Approach

15Proposed Approach

The user profile: a vector of keywords terms corresponding to the user interests implicitly inferred from his activities on social Web systems.

Adjust the importance of each keyword according to the time of its use.

Unified model:

Naturally combine short-term profile and long-term profile into a single.

Give importance to the recent interests without ignoring the continuous ones.

16Main idea

Personalization in this work: weighting the user profile keywords according to the appearing time in addition to the frequency.

Main idea

Revising the notion of frequency by adjusting it with a temporal function

Ensure a unified profile

Illustration17

18Time-Sensitive User Profile

):,...,:,:( 2211Sj

mSj

mSjSjSjSj WtWtWtU ),(.)()( j

cSck

Sck SSKtnTFtW

19Time-Sensitive User Profile

SiDk

kSi

iSi

Sii

tfreq

tfreqtnTF

)(

)()(

².2

)²(exp.

2

1),(

j

C

jC SSSSK

Experiments and Results

21Data Set

01-15 of december , 2013 800 Profiles

69000 Tweets 40 assessors

22Methodology

1. Create the user profile Extract the user tweets

Combine the relative frequencies with the temporal biased function

2. Submit a query to standard search engine Related to the user’s areas of interests defined on Twitter (800

queries)

3. Results Extraction Top 100 Webpages

):,...,:,:( 2211Sj

mSj

mSjSjSjSj WtWtWtU

Methodology

Stop words removal, stemming and tokenization of documents and users’ extracted terms (Apache Lucene classes, Porter Stemming Filter).

3. Create the Webpage-profile Weight according to the tf-idf model

4. Rerank search results

23

),().1(),(.),( QWPSimWPUSimQWPScore

kwpwpwp tttWP ,...,,

21

24Baselines Comparison Results

Standard Search Engine (57.87)

N-tf (62.68)

TSUP (78.15)

P@10

NDCG

Standard Search Engine (45.67)

N-tf (58.80)

TSUP (74.72)

25Impact of User's Profile Information Amount

Influence of the temporal feature: same personalization strategy to compare the time-sensitive user profile (TSUP) with the nTF-based user profile.

Profile Temporal Aspects

Short-term

Long-term

Single

Impact of User's Profile Information Amount

26

27Findings

Promising values:

Term frequency does not reflect the freshness of an interest but gives an overview of how often the user mentioned a term .

Standard search engines return relevant results to the user query’s terms but they are indifferent to the users’ interests

Temporal function: consider the actual interests which are used to enhance the current search without overlooking the persistent interests and helps to personalize recurrent information needs.

28

Conclusion

Conclusion

Problem of personalized search: a user-modeling framework for Twitter microblogging system.

Integration of the social data: accurate and efficient because people are likely to write a blog or bookmark a Webpage about something that interests them.

29

How the temporal-based user profile influences the accuracy of personalized seach using a single profile instead of separately consider the short- and long-term user profiles?

Conclusion

Vector-based representation

Temporal-Frequency

Merging the term frequency

and the freshness of

each keyword(Kernel

Function)

30

Encouraging results: comparison to two non-temporal sensitive approaches.

Aggregation of the current and recurrent interests: increasing amount of information yields to better improvement.

Future Work

Improve experiments: study temporal aspects when enriching the user profile by including diverse user’s social behaviors on the Web.

Comparison with other temporal models

31

Thank You For Your Attention

AMENI [email protected]

PHD STUDENT

PAUL SABATIER UNIVERSITY FRANCE

HIGH INSTITUTE OF MANAGEMENT,TUNISIA